U.S. patent application number 16/843707 was filed with the patent office on 2020-07-23 for motion vector refinement of a motion vector pointing to a fractional sample position.
The applicant listed for this patent is Huawei Technologies Co., Ltd.. Invention is credited to Semih ESENLIK, Han GAO, Anand Meher KOTRA, Zhijie ZHAO.
Application Number | 20200236398 16/843707 |
Document ID | 20200236398 / US20200236398 |
Family ID | 60080795 |
Filed Date | 2020-07-23 |
Patent Application | download [pdf] |
![](/patent/app/20200236398/US20200236398A1-20200723-D00000.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00001.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00002.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00003.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00004.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00005.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00006.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00007.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00008.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00009.png)
![](/patent/app/20200236398/US20200236398A1-20200723-D00010.png)
View All Diagrams
United States Patent
Application |
20200236398 |
Kind Code |
A1 |
ESENLIK; Semih ; et
al. |
July 23, 2020 |
MOTION VECTOR REFINEMENT OF A MOTION VECTOR POINTING TO A
FRACTIONAL SAMPLE POSITION
Abstract
The present disclosure relates to motion vector determination
employing template matching. At first, an initial motion vector is
obtained. If the initial motion vector points to a position which
is fractional and thus, requires interpolation from the integer
sample positions, the initial motion vector is rounded to a closest
integer sample position. The rounded position is then used to
define the search space for the template matching to define a
refinement of the initial motion vector.
Inventors: |
ESENLIK; Semih; (Munich,
DE) ; ZHAO; Zhijie; (Munich, DE) ; GAO;
Han; (Munich, DE) ; KOTRA; Anand Meher;
(Munich, DE) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Huawei Technologies Co., Ltd. |
Shenzhen |
|
CN |
|
|
Family ID: |
60080795 |
Appl. No.: |
16/843707 |
Filed: |
April 8, 2020 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
PCT/EP2017/075714 |
Oct 9, 2017 |
|
|
|
16843707 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04N 19/513 20141101;
H04N 19/52 20141101; H04N 19/533 20141101; H04N 19/57 20141101;
H04N 19/573 20141101; H04N 19/56 20141101 |
International
Class: |
H04N 19/57 20060101
H04N019/57; H04N 19/52 20060101 H04N019/52; H04N 19/533 20060101
H04N019/533; H04N 19/56 20060101 H04N019/56; H04N 19/573 20060101
H04N019/573 |
Claims
1. An apparatus for determination of a motion vector for a
prediction block, the apparatus comprising a processing circuitry
configured to: obtain an initial motion vector and a template for
the prediction block; and determine a refinement of the initial
motion vector by template matching with the template in a search
space, wherein, based upon the initial motion vector pointing to a
fractional sample position, the search space is located on a
position given by rounding the fractional sample position pointed
to by the initial motion vector to an integer sample position.
2. The apparatus according to claim 1, the processing circuitry
being further configured to: calculate a template matching cost for
the initial motion vector pointing to the fractional sample
position, compare the calculated template matching cost of the
fractional sample position with a template matching cost of the
determined refinement of the initial motion vector, and determine
the motion vector for the prediction block based on the fractional
sample position upon a state where the template matching cost of
the determined refinement is not lower than the calculated template
matching cost of the fractional sample position.
3. The apparatus according to claim 1, wherein the search space
comprises of sample positions spaced by integer sample step from
each other.
4. The apparatus according to claim 1, wherein the search space is
formed by K rows and L columns of samples located in rows and
columns in integer distance from each other, wherein the rounded
fractional position is one of the samples of the search space.
5. The apparatus according to claim 1, wherein the rounded
fractional position is located substantially in a center of the
search space.
6. The apparatus according to claim 1, wherein the processing
circuitry is configured to perform the template matching for the
search space of which the sample positions are determined in a
plurality of steps including: determining an integer search space
comprising of sample positions spaced by integer sample step from
each other, performing template matching in the integer search
space to obtain a best matching integer position, based on the best
matching integer position, determining a fractional search space
with at least one sample spaced by less than integer sample step
from the closest integer search space sample positions, and
performing template matching in the fractional search space to
obtain a best matching position.
7. The apparatus according to claim 1, wherein the rounding is
obtained as an integer sample position closest to the fractional
sample position to which the initial motion vector points.
8. The apparatus according to claim 7, wherein, upon a case where
there are more than one integer sample position equally close to
the fractional sample position, the rounding is obtained as an
integer sample position in a predefined direction, the predefined
direction being left, right, top, or bottom.
9. The apparatus according to claim 7, wherein, upon a case where
there are more than one integer sample positions equally close to
the fractional sample position, the search space is a unity of a
search space located on the more than one integer sample positions
equally close to the fractional sample position.
10. The apparatus according to claim 7, wherein, upon a case where
there are more than one integer sample positions equally close to
the fractional sample position, the rounding is obtained as the one
among the equally close integer sample positions which results in a
shorter magnitude of the rounded initial motion vector.
11. The apparatus according to claim 1, wherein: the search space
is located on a position given by the rounded initial motion vector
and comprises one or more fractional sample positions, each of the
fractional sample positions belonging to the search space is
obtained by interpolation filtering with a filter of a predefined
tap-size assessing integer samples only within a window, and the
window is formed by integer samples accessible for the template
matching in the search space.
12. The apparatus according to claim 1, wherein: the search space
is located on a position given by the initial motion vector, and
the template matching: accesses integer sample positions within a
predetermined window, the predetermined window comprising the
search space and further integer sample positions accessible for
the template matching, replaces at least one of the integer sample
positions located out of the window by a replacement value that is
based on at least one sample within the window, and uses the
accessed integer sample position values as well as the replacement
integer sample position values to perform the template
matching.
13. A prediction block encoding apparatus for encoding a prediction
block comprising: the apparatus for determination of the motion
vector for the prediction block according to claim 1; and an
encoder configured to encode differences between the prediction
block and a predictor obtained according to the motion vector,
resulting in a bitstream.
14. A decoding apparatus for decoding the prediction block from a
bitstream comprising: a decoder configured to decode from the
bitstream differences between the prediction block and a predictor
obtained according to the motion vector; the apparatus for
determination of the motion vector for the prediction block
according to claim 1; and a motion predictor configured to
determine the predictor according to the motion vector; and a
reconstructor configured to reconstruct the prediction block from
the predictor and the decoded differences.
15. A method for determination of a motion vector for a prediction
block, the method comprising: obtaining an initial motion vector
and a template for the prediction block; and determining a
refinement of the initial motion vector by template matching with
the template in a search space, wherein, based upon the initial
motion vector pointing to a fractional sample position, the search
space is located on a position given by rounding the fractional
sample position pointed to by the initial motion vector to an
integer sample position.
16. A method for encoding the prediction block, the method
comprising: determining of the motion vector for the prediction
block according to the method of claim 15; and encoding differences
between the prediction block and a predictor obtained according to
the motion vector, resulting in a bitstream.
17. A method for decoding the prediction block from a bitstream,
the method comprising: decoding from the bitstream differences
between the prediction block and a predictor obtained according to
the motion vector; determining of the motion vector for the
prediction block according to the method of claim 15; determining
the predictor according to the motion vector; and reconstructing
the prediction block from the predictor and the decoded
differences.
18. A non-transitory computer readable medium storing instructions
which when executed on a processor cause the processor to perform
the method according to claim 15.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of International
Application No. PCT/EP2017/075714, filed on Oct. 9, 2017. The
aforementioned application is hereby incorporated by reference in
its entirety.
TECHNICAL FIELD
[0002] The present invention relates to motion vector determination
and refinement which may be employed during encoding and decoding
of videos.
BACKGROUND
[0003] Current hybrid video codecs employ predictive coding. A
picture of a video sequence is subdivided into blocks of pixels and
these blocks are then coded. Instead of coding a block pixel by
pixel, the entire block is predicted using already encoded pixels
in the spatial or temporal proximity of the block. The encoder
further processes only the differences between the block and its
prediction. The further processing typically includes a
transformation of the block pixels into coefficients in a
transformation domain. The coefficients may then be further
compressed by means of quantization and further compacted by
entropy coding to form a bitstream. The bitstream further includes
any signaling information which enables the decoder to decode the
encoded video. For instance, the signaling may include settings
concerning the encoder settings such as size of the input picture,
frame rate, quantization step indication, prediction applied to the
blocks of the pictures, or the like.
[0004] Temporal prediction exploits temporal correlation between
pictures, also referred to as frames, of a video. The temporal
prediction is also called inter-prediction, as it is a prediction
using the dependencies between (inter) different video frames.
Accordingly, a block being encoded, also referred to as a current
block, is predicted from one or more previously encoded picture(s)
referred to as a reference picture(s). A reference picture is not
necessarily a picture preceding the current picture in which the
current block is located in the displaying order of the video
sequence. The encoder may encode the pictures in a coding order
different from the displaying order. As a prediction of the current
block, a co-located block in a reference picture may be determined.
The co-located block is a block which is located in the reference
picture on the same position as is the current block in the current
picture. Such prediction is accurate for motionless picture
regions, i.e. picture regions without movement from one picture to
another.
[0005] In order to obtain a predictor which takes into account the
movement, i.e. a motion compensated predictor, motion estimation is
typically employed when determining the prediction of the current
block. Accordingly, the current block is predicted by a block in
the reference picture, which is located in a distance given by a
motion vector from the position of the co-located block. In order
to enable a decoder to determine the same prediction of the current
block, the motion vector may be signaled in the bitstream. In order
to further reduce the signaling overhead caused by signaling the
motion vector for each of the blocks, the motion vector itself may
be estimated. The motion vector estimation may be performed based
on the motion vectors of the neighboring blocks in spatial and/or
temporal domain.
[0006] The prediction of the current block may be computed using
one reference picture or by weighting predictions obtained from two
or more reference pictures. The reference picture may be an
adjacent picture, i.e. a picture immediately preceding and/or the
picture immediately following the current picture in the display
order since adjacent pictures are most likely to be similar to the
current picture. However, in general, the reference picture may be
also any other picture preceding or following the current picture
in the displaying order and preceding the current picture in the
bitstream (decoding order). This may provide advantages for
instance in case of occlusions and/or non-linear movement in the
video content. The reference picture identification may thus be
also signaled in the bitstream.
[0007] A special mode of the inter-prediction is a so-called
bi-prediction in which two reference pictures are used in
generating the prediction of the current block. In particular, two
predictions determined in the respective two reference pictures are
combined into a prediction signal of the current block. The
bi-prediction may result in a more accurate prediction of the
current block than the uni-prediction, i.e. prediction only using a
single reference picture. The more accurate prediction leads to
smaller differences between the pixels of the current block and the
prediction (referred to also as "residuals"), which may be encoded
more efficiently, i.e. compressed to a shorter bitstream. In
general, more than two reference pictures may be used to find
respective more than two reference blocks to predict the current
block, i.e. a multi-reference inter prediction can be applied. The
term multi-reference prediction thus includes bi-prediction as well
as predictions using more than two reference pictures.
[0008] In order to provide more accurate motion estimation, the
resolution of the reference picture may be enhanced by
interpolating samples between pixels. Fractional pixel
interpolation can be performed by weighted averaging of the closest
pixels. In case of half-pixel resolution, for instance a bilinear
interpolation is typically used. Other fractional pixels are
calculated as an average of the closest pixels weighted by the
inverse of the distance between the respective closest pixels to
the pixel being predicted.
[0009] The motion vector estimation is a computationally complex
task in which a similarity is calculated between the current block
and the corresponding prediction blocks pointed to by candidate
motion vectors in the reference picture. Typically, the search
region includes M.times.M samples of the image and each of the
sample position of the M.times.M candidate positions is tested. The
test includes calculation of a similarity measure between the
N.times.N reference block C and a block R, located at the tested
candidate position of the search region. For its simplicity, the
sum of absolute differences (SAD) is a measure frequently used for
this purpose and given by:
SAD ( x , y ) = i = 0 N - 1 j = 0 N - 1 R i , j ( x , y ) - C i , j
##EQU00001##
[0010] In the above formula, x and y define the candidate position
within the search region, while indices i and j denote samples
within the reference block C and candidate block R. The candidate
position is often referred to as block displacement or offset,
which reflects the representation of the block matching as shifting
of the reference block within the search region and calculating a
similarity between the reference block C and the overlapped portion
of the search region. In order to reduce the complexity, the number
of candidate motion vectors is usually reduced by limiting the
candidate motion vectors to a certain search space. The search
space may be, for instance, defined by a number and/or positions of
pixels surrounding the position in the reference picture
corresponding to the position of the current block in the current
image. After calculating SAD for all M.times.M candidate positions
x and y, the best matching block R is the block on the position
resulting in the lowest SAD, corresponding to the largest
similarity with reference block C. On the other hand, the candidate
motion vectors may be defined by a list of candidate motion vectors
formed by motion vectors of neighboring blocks.
[0011] Motion vectors are usually at least partially determined at
the encoder side and signaled to the decoder within the coded
bitstream. However, the motion vectors may also be derived at the
decoder. In such case, the current block is not available at the
decoder and cannot be used for calculating the similarity to the
blocks to which the candidate motion vectors point in the reference
picture. Therefore, instead of the current block, a template is
used which is constructed out of pixels of already decoded blocks.
For instance, already decoded pixels adjacent to the current block
may be used. Such motion estimation provides an advantage of
reducing the signaling: the motion vector is derived in the same
way at both the encoder and the decoder and thus, no signaling is
needed. On the other hand, the accuracy of such motion estimation
may be lower.
[0012] In order to provide a tradeoff between the accuracy and
signaling overhead, the motion vector estimation may be divided
into two steps: motion vector derivation and motion vector
refinement. For instance, a motion vector derivation may include
selection of a motion vector from the list of candidates. Such a
selected motion vector may be further refined for instance by a
search within a search space. The search in the search space is
based on calculating cost function for each candidate motion
vector, i.e. for each candidate position of block to which the
candidate motion vector points.
[0013] Document JVET-D0029: Decoder-Side Motion Vector Refinement
Based on Bilateral Template Matching, X. Chen, J. An, J. Zheng (the
document can be found at: http://phenix.it-sudparis.eu/jvet/site)
shows motion vector refinement in which a first motion vector in
integer pixel resolution is found and further refined by a search
with a half-pixel resolution in a search space around the first
motion vector.
[0014] In order to perform motion vector refinement, it is
necessary to store at least those samples in the memory, which are
necessary for the current block to perform the refinement, i.e. the
samples which correspond to the search space and samples which can
be accessed when template matching in the search space is
performed.
[0015] External memory access is an important design parameter in
present hardware architectures and/or software implementations.
This is caused by the fact that the external memory access slows
down the processing in comparison with the intern memory
utilization. On the other hand, internal memory on chip is limited,
for instance due to the chip size implementation.
SUMMARY
[0016] The present disclosure is based on observation that motion
vector refinement when implemented in combination with fractional
interpolation may require further increase of on-chip memory size
or even external memory access. Both options may be undesirable.
Moreover, when the initial motion vector points to a fractional
position, i.e. a position with fractional coordinate, the search
space including sample in 1-sample or multiples of 1-sample
distance from each other and the initial motion vector result in
fractional positions which are to be all interpolated.
[0017] In view of the above mentioned problem, the present
disclosure provides motion vector prediction which enables to take
into account the number of accesses to the external memory and the
number of samples which are necessary to be accessible for motion
vector refinement of a motion vector for a coding block. This may
be achieved by rounding the initial motion vector coordinates in
case they are fractional.
[0018] According to an aspect of the invention, an apparatus is
provided for determination of a motion vector for a prediction
block including a processing circuitry configured to: obtain an
initial motion vector and a template for the prediction block;
determine a refinement of the initial motion vector by template
matching with said template in a search space. When the initial
motion vector points to a fractional sample position, the search
space is located on a position given by rounding the fractional
position pointed to by the initial motion vector to an integer
sample position.
[0019] Such apparatus may provide an advantage of reducing the
number of fractional positions necessary for the template matching
which on one hand reduces the complexity since less interpolation
operations are necessary and, on the other hand, reduces the number
of additional integer samples necessary for the memory access
window, since the number of fractional positions and thus also
integer sample positions necessary for their interpolation is
reduced.
[0020] In one embodiment, the processing circuitry is further
configured for calculating the template matching cost for the
initial motion vector pointing to the fractional sample position;
comparing the calculated template matching cost of said fractional
sample position with the template matching cost of the determined
refinement of the initial motion vector; and determining the motion
vector for the prediction block based on said fractional sample
position if the template matching cost of the determined refinement
is not lower than the calculated template matching cost of said
fractional sample position.
[0021] For example, the search space consists of sample positions
spaced by integer sample step from each other. The search space is
formed, for instance, by K rows and L columns of samples located in
rows and columns in integer distance from each other, wherein the
rounded fractional position is one of the samples of the search
space.
[0022] According to an example, the rounded fractional position is
located substantially in the center of the search space.
[0023] The processing circuitry can be further configured to
perform the template matching for a search space of which the
sample positions are determined in a plurality of steps including:
determining an integer search space consisting of sample positions
spaced by integer sample step from each other; performing template
matching in the integer search space to obtain a best matching
integer position; based on the best matching integer position,
determining a fractional search space with at least one sample
spaced by less than integer sample step from the closest integer
search space sample positions; and performing template matching in
the fractional search space to obtain a best matching position.
[0024] In one example, the rounding is obtained as an integer
sample position closest to the fractional position to which the
initial motion vector points.
[0025] If there are more than one integer sample position equally
close to the fractional position, in one example, the rounding is
obtained as an integer sample position in a predefined direction,
the predefined direction being left, right, top or bottom.
[0026] If there are more than one integer sample positions equally
close to the fractional position, in another example, the search
space is a unity of a search space located on said more than one
integer sample positions equally close to the fractional
position.
[0027] If there are more than one integer sample positions equally
close to the fractional position, in yet another example, the
rounding is obtained as the one among the equally close integer
sample positions which results in a shorter magnitude of the
rounded initial motion vector.
[0028] According to an aspect of the invention, an apparatus is
provided for encoding a prediction block including: the apparatus
for determination of a motion vector for the prediction block as
described above; and an encoding unit for encoding differences
between the prediction block and a predictor obtained according to
the motion vector, resulting in a bitstream.
[0029] According to another aspect of the invention, an apparatus
is provided for decoding a prediction block from a bitstream
including: a decoding unit for decoding from the bitstream
differences between the prediction block and a predictor obtained
according to a motion vector; the apparatus for determination of
the motion vector for the prediction block as described above; a
motion prediction unit for determining the predictor according to
the motion vector; and a reconstructor for reconstructing the
prediction block from the predictor and the decoded
differences.
[0030] According to an aspect of the invention, a method is
provided for determination of a motion vector for a prediction
block including the steps of: obtaining an initial motion vector
and a template for the prediction block; determining a refinement
of the initial motion vector by template matching with said
template in a search space; wherein, when the initial motion vector
points to a fractional sample position, the search space is located
on a position given by rounding the fractional position pointed to
by the initial motion vector to an integer sample position.
[0031] In one embodiment, the method further comprising the steps
of: calculating the template matching cost for the initial motion
vector pointing to the fractional sample position; comparing the
calculated template matching cost of said fractional sample
position with the template matching cost of the determined
refinement of the initial motion vector; and determining the motion
vector for the prediction block based on said fractional sample
position if the template matching cost of the determined refinement
is not lower than the calculated template matching cost of said
fractional sample position.
[0032] In an exemplary implementation, the search space consists of
sample positions spaced by integer sample step from each other.
[0033] For instance, the search space is formed by K rows and L
columns of samples located in rows and columns in integer distance
from each other, wherein the rounded fractional position is one of
the samples of the search space. K and L here are non-zero integers
of which at least one is larger than 1.
[0034] The rounded fractional position may be located substantially
in the center of the search space.
[0035] In one embodiment, which may be combined with any of the
above embodiments and examples, the template matching is performed
for a search space of which the sample positions are determined in
a plurality of steps including: determining an integer search space
consisting of sample positions spaced by integer sample step from
each other; performing template matching in the integer search
space to obtain a best matching integer position; based on the best
matching integer position, determining a fractional search space
with at least one sample spaced by less than integer sample step
from the closest integer search space sample positions; and
performing template matching in the fractional search space to
obtain a best matching position.
[0036] In addition or alternatively to the above embodiments and
examples, the rounding is obtained as an integer sample position
closest to the fractional position to which the initial motion
vector points.
[0037] In one example, if there are more than one integer sample
position equally close to the fractional position, the rounding is
obtained as an integer sample position in a predefined direction,
the predefined direction being left, right, top or bottom.
[0038] In another example, if there are more than one integer
sample positions equally close to the fractional position, the
search space is a unity of a search space located on said more than
one integer sample positions equally close to the fractional
position.
[0039] In still another example, if there are more than one integer
sample positions equally close to the fractional position, the
rounding is obtained as the one among the equally close integer
sample positions which results in a shorter magnitude of the
rounded initial motion vector.
[0040] According to an aspect of the invention, a method is
provided for encoding a prediction block including the steps of
determining of a motion vector for the prediction block according
to any of the above described embodiments and examples; and
encoding differences between the prediction block and a predictor
obtained according to the motion vector, resulting in a
bitstream.
[0041] According to an aspect of the invention, a method is
provided for decoding a prediction block from a bitstream including
the steps of: decoding from the bitstream differences between the
prediction block and a predictor obtained according to a motion
vector; determining of the motion vector for the prediction block
according to any of the above described embodiments and examples;
determining the predictor according to the motion vector; and
reconstructing the prediction block from the predictor and the
decoded differences.
[0042] According to an aspect of the invention a non-transitory
computer-readable storage medium is provided storing instructions
which when executed by a processor/processing circuitry perform the
steps according to any of the above aspects or embodiments or their
combinations.
BRIEF DESCRIPTION OF DRAWINGS
[0043] In the following exemplary embodiments are described in more
detail with reference to the attached figures and drawings, in
which:
[0044] FIG. 1 is a block diagram showing an exemplary structure of
an encoder in which the motion vector derivation and refinement may
be employed;
[0045] FIG. 2 is a block diagram showing an exemplary structure of
a decoder in which the motion vector derivation and refinement may
be employed;
[0046] FIG. 3 is a schematic drawing illustrating an exemplary
template matching suitable for bi-prediction;
[0047] FIG. 4 is a schematic drawing illustrating an exemplary
template matching suitable for uni- and bi-prediction;
[0048] FIG. 5 is a block diagram illustrating stages of motion
vector derivation operating without providing initial motion
vectors to be refined in the bitstream;
[0049] FIG. 6 is a block diagram illustrating an exemplary hardware
to implement an embodiment of the invention;
[0050] FIG. 7 is a schematic drawing illustrating for a coding
block an exemplary window with samples which should be available to
be accessed;
[0051] FIG. 8 is a schematic drawing illustrating iterative search
space;
[0052] FIG. 9 is a schematic drawing illustrating extension of the
memory access window in horizontal direction due to interpolation
filtering;
[0053] FIG. 10 is a schematic drawing illustrating adjustment of a
fractional initial motion vector;
[0054] FIG. 11 is a schematic drawing illustrating definition of a
sub-window for fractional sample positions;
[0055] FIG. 12 is a schematic drawing illustrating comparison
between the search space obtained without and with rounding of the
initial motion vector coordinates;
[0056] FIG. 13A is a flow diagram illustrating motion vector
refinement without rounding of the initial motion vector;
[0057] FIG. 13B is a flow diagram illustrating motion vector
refinement with rounding of the initial motion vector to a
predetermined integer position;
[0058] FIG. 13C is a flow diagram illustrating motion vector
refinement without rounding of the initial motion vector to a
plurality of integer positions;
[0059] FIG. 14 is a schematic drawing illustrating definition of a
sub-window for fractional sample positions;
[0060] FIG. 15 is a schematic drawing illustrating exemplary
positions including fractional positions which are allowed to form
search space positions for motion vector refinement;
[0061] FIG. 16 is a schematic drawing illustrating exemplary
fractional positions which are not allowed to form search space
positions for motion vector refinement;
[0062] FIG. 17 is a flow diagram illustrating method for
determining which position of a search space are allowed to be
tested by template matching for motion vector refinement
[0063] FIG. 18 is a schematic drawing illustrating padding of two
samples to enable fractional position calculation;
[0064] FIG. 19 is a schematic drawing illustrating padding of four
samples to enable template matching on the search space boundaries;
and
[0065] FIG. 20 is a flow chart of a method applying sample
padding.
DETAILED DESCRIPTION
[0066] As mentioned above, the external memory access is one of the
most important design considerations in today's hardware and
software architectures. Motion vector estimation especially when
including template matching, for instance in case of motion vector
refinement, may also be used with interpolation filtering to obtain
fractional positions of the search space. Use of the interpolation
filtering may require increase of the number of samples which need
to be accessed from the memory. However, this may lead to either
increase of the expensive on-chip memory or to increased number of
external memory accesses, which on the other hand slows down the
implementation. Especially at the decoder side, these problems may
lead to more expensive or slower applications which is not
desirable.
[0067] According to the present disclosure, in order to limit the
memory access window, the interpolation of the fractional positions
is reduced or avoided by determining the search space relatively to
the position pointed to by the initial motion vector shifted to the
closest integer position.
[0068] According to an embodiment of the invention an apparatus is
provided for determination of a motion vector for a prediction
block including a processing circuitry. The processing circuitry is
configured to obtain an initial motion vector and a template for
the prediction block and to determine a refinement of the initial
motion vector by template matching with said template in a search
space. When the initial motion vector points to a fractional sample
position, the search space is located on a position given by
rounding the fractional position pointed to by the initial motion
vector to an integer sample position.
[0069] FIG. 1 shows an encoder 100 which comprises an input for
receiving input image samples of frames or pictures of a video
stream and an output for generating an encoded video bitstream. The
term "frame" in this disclosure is used as a synonym for picture.
However, it is noted that the present disclosure is also applicable
to fields in case interlacing is applied. In general, a picture
includes m times n pixels. This corresponds to image samples and
may comprise one or more color components. For the sake of
simplicity, the following description refers to pixels meaning
samples of luminance. However, it is noted that the motion vector
search of the invention can be applied to any color component
including chrominance or components of a search space such as RGB
or the like. On the other hand, it may be beneficial to only
perform motion vector estimation for one component and to apply the
determined motion vector to more (or all) components.
[0070] The input blocks to be coded do not necessarily have the
same size. One picture may include blocks of different sizes and
the block raster of different pictures may also differ.
[0071] In an explicative realization, the encoder 100 is configured
to apply prediction, transformation, quantization, and entropy
coding to the video stream. The transformation, quantization, and
entropy coding are carried out respectively by a transform unit
106, a quantization unit 108 and an entropy encoding unit 170 so as
to generate as an output the encoded video bitstream.
[0072] The video stream may include a plurality of frames, wherein
each frame is divided into blocks of a certain size that are either
intra or inter coded. The blocks of for example the first frame of
the video stream are intra coded by means of an intra prediction
unit 154. An intra frame is coded using only the information within
the same frame, so that it can be independently decoded and it can
provide an entry point in the bitstream for random access. Blocks
of other frames of the video stream may be inter coded by means of
an inter prediction unit 144: information from previously coded
frames (reference frames) is used to reduce the temporal
redundancy, so that each block of an inter-coded frame is predicted
from a block in a reference frame. A mode selection unit 160 is
configured to select whether a block of a frame is to be processed
by the intra prediction unit 154 or the inter prediction unit 144.
This mode selection unit 160 also controls the parameters of intra
or inter prediction. In order to enable refreshing of the image
information, intra-coded blocks may be provided within inter-coded
frames. Moreover, intra-frames which contain only intra-coded
blocks may be regularly inserted into the video sequence in order
to provide entry points for decoding, i.e. points where the decoder
can start decoding without having information from the previously
coded frames.
[0073] The intra estimation unit 152 and the intra prediction unit
154 are units which perform the intra prediction. In particular,
the intra estimation unit 152 may derive the prediction mode based
also on the knowledge of the original image while intra prediction
unit 154 provides the corresponding predictor, i.e. samples
predicted using the selected prediction mode, for the difference
coding. For performing spatial or temporal prediction, the coded
blocks may be further processed by an inverse quantization unit
110, and an inverse transform unit 112. After reconstruction of the
block a loop filtering unit 120 is applied to further improve the
quality of the decoded image. The filtered blocks then form the
reference frames that are then stored in a decoded picture buffer
130. Such decoding loop (decoder) at the encoder side provides the
advantage of producing reference frames which are the same as the
reference pictures reconstructed at the decoder side. Accordingly,
the encoder and decoder side operate in a corresponding manner. The
term "reconstruction" here refers to obtaining the reconstructed
block by adding to the decoded residual block the prediction
block.
[0074] The inter estimation unit 142 receives as an input a block
of a current frame or picture to be inter coded and one or several
reference frames from the decoded picture buffer 130. Motion
estimation is performed by the inter estimation unit 142 whereas
motion compensation is applied by the inter prediction unit 144.
The motion estimation is used to obtain a motion vector and a
reference frame based on certain cost function, for instance using
also the original image to be coded. For example, the motion
estimation unit 142 may provide initial motion vector estimation.
The initial motion vector may then be signaled within the bitstream
in form of the vector directly or as an index referring to a motion
vector candidate within a list of candidates constructed based on a
predetermined rule in the same way at the encoder and the decoder.
The motion compensation then derives a predictor of the current
block as a translation of a block co-located with the current block
in the reference frame to the reference block in the reference
frame, i.e. by a motion vector. The inter prediction unit 144
outputs the prediction block for the current block, wherein said
prediction block minimizes the cost function. For instance, the
cost function may be a difference between the current block to be
coded and its prediction block, i.e. the cost function minimizes
the residual block. The minimization of the residual block is based
e.g. on calculating a sum of absolute differences (SAD) between all
pixels (samples) of the current block and the candidate block in
the candidate reference picture. However, in general, any other
similarity metric may be employed, such as mean square error (MSE)
or structural similarity metric (SSIM).
[0075] However, cost-function may also be the number of bits
necessary to code such inter-block and/or distortion resulting from
such coding. Thus, the rate-distortion optimization procedure may
be used to decide on the motion vector selection and/or in general
on the encoding parameters such as whether to use inter or intra
prediction for a block and with which settings.
[0076] The intra estimation unit 152 and inter prediction unit 154
receive as an input a block of a current frame or picture to be
intra coded and one or several reference samples from an already
reconstructed area of the current frame. The intra prediction then
describes pixels of a current block of the current frame in terms
of a function of reference samples of the current frame. The intra
prediction unit 154 outputs a prediction block for the current
block, wherein said prediction block advantageously minimizes the
difference between the current block to be coded and its prediction
block, i.e., it minimizes the residual block. The minimization of
the residual block can be based e.g. on a rate-distortion
optimization procedure. In particular, the prediction block is
obtained as a directional interpolation of the reference samples.
The direction may be determined by the rate-distortion optimization
and/or by calculating a similarity measure as mentioned above in
connection with inter-prediction.
[0077] The inter estimation unit 142 receives as an input a block
or a more universal-formed image sample of a current frame or
picture to be inter coded and two or more already decoded pictures
231. The inter prediction then describes a current image sample of
the current frame in terms of motion vectors to reference image
samples of the reference pictures. The inter prediction unit 142
outputs one or more motion vectors for the current image sample,
wherein said reference image samples pointed to by the motion
vectors advantageously minimize the difference between the current
image sample to be coded and its reference image samples, i.e., it
minimizes the residual image sample. The predictor for the current
block is then provided by the inter prediction unit 144 for the
difference coding.
[0078] The difference between the current block and its prediction,
i.e. the residual block 105, is then transformed by the transform
unit 106. The transform coefficients 107 are quantized by the
quantization unit 108 and entropy coded by the entropy encoding
unit 170. The thus generated encoded picture data 171, i.e. encoded
video bitstream, comprises intra coded blocks and inter coded
blocks and the corresponding signaling (such as the mode
indication, indication of the motion vector, and/or
intra-prediction direction). The transform unit 106 may apply a
linear transformation such as a Fourier or Discrete Cosine
Transformation (DFT/FFT or DCT). Such transformation into the
spatial frequency domain provides the advantage that the resulting
coefficients 107 have typically higher values in the lower
frequencies. Thus, after an effective coefficient scanning (such as
zig-zag), and quantization, the resulting sequence of values has
typically some larger values at the beginning and ends with a run
of zeros. This enables further efficient coding. Quantization unit
108 performs the actual lossy compression by reducing the
resolution of the coefficient values. The entropy coding unit 170
then assigns to coefficient values binary codewords to produce a
bitstream. The entropy coding unit 170 also codes the signaling
information (not shown in FIG. 1).
[0079] FIG. 2 shows a video decoder 200. The video decoder 200
comprises particularly a decoded picture buffer 230, an inter
prediction unit 244 and an intra prediction unit 254, which is a
block prediction unit. The decoded picture buffer 230 is configured
to store at least one (for uni-prediction) or at least two (for
bi-prediction) reference frames reconstructed from the encoded
video bitstream, said reference frames being different from a
current frame (currently decoded frame) of the encoded video
bitstream. The intra prediction unit 254 is configured to generate
a prediction block, which is an estimate of the block to be
decoded. The intra prediction unit 254 is configured to generate
this prediction based on reference samples that are obtained from
the decoded picture buffer 230.
[0080] The decoder 200 is configured to decode the encoded video
bitstream generated by the video encoder 100, and preferably both
the decoder 200 and the encoder 100 generate identical predictions
for the respective block to be encoded/decoded. The features of the
decoded picture buffer 230 and the intra prediction unit 254 are
similar to the features of the decoded picture buffer 130 and the
intra prediction unit 154 of FIG. 1.
[0081] The video decoder 200 comprises further units that are also
present in the video encoder 100 like e.g. an inverse quantization
unit 210, an inverse transform unit 212, and a loop filtering unit
220, which respectively correspond to the inverse quantization unit
110, the inverse transform unit 112, and the loop filtering unit
120 of the video coder 100.
[0082] An entropy decoding unit 204 is configured to decode the
received encoded video bitstream and to correspondingly obtain
quantized residual transform coefficients 209 and signaling
information. The quantized residual transform coefficients 209 are
fed to the inverse quantization unit 210 and an inverse transform
unit 212 to generate a residual block. The residual block is added
to a prediction block 265 and the addition is fed to the loop
filtering unit 220 to obtain the decoded video. Frames of the
decoded video can be stored in the decoded picture buffer 230 and
serve as a decoded picture 231 for inter prediction.
[0083] Generally, the intra prediction units 154 and 254 of FIGS. 1
and 2 can use reference samples from an already encoded area to
generate prediction signals for blocks that need to be encoded or
need to be decoded.
[0084] The entropy decoding unit 204 receives as its input the
encoded bitstream 171. In general, the bitstream is at first
parsed, i.e. the signaling parameters and the residuals are
extracted from the bitstream. Typically, the syntax and semantic of
the bitstream is defined by a standard so that the encoders and
decoders may work in an interoperable manner. As described in the
above Background section, the encoded bitstream does not only
include the prediction residuals. In case of motion compensated
prediction, a motion vector indication is also coded in the
bitstream and parsed therefrom at the decoder. The motion vector
indication may be given by means of a reference picture in which
the motion vector is provided and by means of the motion vector
coordinates. So far, coding the complete motion vectors was
considered. However, also only the difference between the current
motion vector and the previous motion vector in the bitstream may
be encoded. This approach allows exploiting the redundancy between
motion vectors of neighboring blocks.
[0085] In order to efficiently code the reference picture, H.265
codec (ITU-T, H265, Series H: Audiovisual and multimedia systems:
High Efficient Video Coding) provides a list of reference pictures
assigning to list indices respective reference frames. The
reference frame is then signaled in the bitstream by including
therein the corresponding assigned list index. Such list may be
defined in the standard or signaled at the beginning of the video
or a set of a number of frames. It is noted that in H.265 there are
two lists of reference pictures defined, called L0 and L1. The
reference picture is then signaled in the bitstream by indicating
the list (L0 or L1) and indicating an index in that list associated
with the desired reference picture. Providing two or more lists may
have advantages for better compression. For instance, L0 may be
used for both uni-directionally inter-predicted slices and
bi-directionally inter-predicted slices while L1 may only be used
for bi-directionally inter-predicted slices. However, in general
the present disclosure is not limited to any content of the L0 and
L1 lists.
[0086] The lists L0 and L1 may be defined in the standard and
fixed. However, more flexibility in coding/decoding may be achieved
by signaling them at the beginning of the video sequence.
Accordingly, the encoder may configure the lists L0 and L1 with
particular reference pictures ordered according to the index. The
L0 and L1 lists may have the same fixed size. There may be more
than two lists in general. The motion vector may be signaled
directly by the coordinates in the reference picture.
Alternatively, as also specified in H.265, a list of candidate
motion vectors may be constructed and an index associated in the
list with the particular motion vector can be transmitted.
[0087] Motion vectors of the current block are usually correlated
with the motion vectors of neighboring blocks in the current
picture or in the earlier coded pictures. This is because
neighboring blocks are likely to correspond to the same moving
object with similar motion and the motion of the object is not
likely to change abruptly over time. Consequently, using the motion
vectors in neighboring blocks as predictors reduces the size of the
signaled motion vector difference. The Motion Vector Predictors
(MVPs) are usually derived from already encoded/decoded motion
vectors from spatial neighboring blocks or from temporally
neighboring blocks in the co-located picture. In H.264/AVC, this is
done by doing a component wise median of three spatially
neighboring motion vectors. Using this approach, no signaling of
the predictor is required. Temporal MVPs from a co-located picture
are only considered in the so called temporal direct mode of
H.264/AVC. The H.264/AVC direct modes are also used to derive other
motion data than the motion vectors. Hence, they relate more to the
block merging concept in HEVC. In HEVC, the approach of implicitly
deriving the MVP was replaced by a technique known as motion vector
competition, which explicitly signals which MVP from a list of
MVPs, is used for motion vector derivation. The variable coding
quad-tree block structure in HEVC can result in one block having
several neighboring blocks with motion vectors as potential MVP
candidates. Taking the left neighbor as an example, in the worst
case a 64.times.64 luma prediction block could have 16 4.times.4
luma prediction blocks to the left when a 64.times.64 luma coding
tree block is not further split and the left one is split to the
maximum depth.
[0088] Advanced Motion Vector Prediction (AMVP) was introduced to
modify motion vector competition to account for such a flexible
block structure. During the development of HEVC, the initial AMVP
design was significantly simplified to provide a good trade-off
between coding efficiency and an implementation friendly design.
The initial design of AMVP included five MVPs from three different
classes of predictors: three motion vectors from spatial neighbors,
the median of the three spatial predictors and a scaled motion
vector from a co-located, temporally neighboring block.
Furthermore, the list of predictors was modified by reordering to
place the most probable motion predictor in the first position and
by removing redundant candidates to assure minimal signaling
overhead. The final design of the AMVP candidate list construction
includes the following two MVP candidates: a) up to two spatial
candidate MVPs that are derived from five spatial neighboring
blocks; b) one temporal candidate MVPs derived from two temporal,
co-located blocks when both spatial candidate MVPs are not
available or they are identical; and c) zero motion vectors when
the spatial, the temporal or both candidates are not available.
Details on motion vector determination can be found in the book by
V. Sze et al (Ed.), High Efficiency Video Coding (HEVC): Algorithms
and Architectures, Springer, 2014, in particular in Chapter 5,
incorporated herein by reference.
[0089] In order to further improve motion vector estimation without
further increase in signaling overhead, it may be beneficial to
further refine the motion vectors derived at the encoder side and
provided in the bitstream. The motion vector refinement may be
performed at the decoder without assistance from the encoder. The
encoder in its decoder loop may employ the same refinement to
obtain corresponding motion vectors. Motion vector refinement is
performed in a search space which includes integer pixel positions
and fractional pixel positions of a reference picture. For example,
the fractional pixel positions may be half-pixel positions or
quarter-pixel or further fractional positions. The fractional pixel
positions may be obtained from the integer (full-pixel) positions
by interpolation such as bi-linear interpolation.
[0090] In a bi-prediction of current block, two prediction blocks
obtained using the respective first motion vector of list L0 and
the second motion vector of list L1, are combined to a single
prediction signal, which can provide a better adaptation to the
original signal than uni-prediction, resulting in less residual
information and possibly a more efficient compression.
[0091] Since at the decoder, the current block is not available
since it is being decoded, for the purpose of motion vector
refinement, a template is used, which is an estimate of the current
block and which is constructed based on the already processed (i.e.
coded at the encoder side and decoded at the decoder side) image
portions.
[0092] First, an estimate of the first motion vector MV0 and an
estimate of the second motion vector MV1 are received as input at
the decoder 200. At the encoder side 100, the motion vector
estimates MV0 and MV1 may be obtained by block matching and/or by
search in a list of candidates (such as merge list) formed by
motion vectors of the blocks neighboring to the current block (in
the same picture or in adjacent pictures). MV0 and MV1 are then
advantageously signaled to the decoder side within the bitstream.
However, it is noted that in general, also the first determination
stage at the encoder could be performed by template matching which
would provide the advantage of reducing signaling overhead.
[0093] At the decoder side 200, the motion vectors MV0 and MV1 are
advantageously obtained based on information in the bitstream. The
MV0 and MV1 are either directly signaled, or differentially
signaled, and/or an index in the list of motion vector (merge list)
is signaled. However, the present disclosure is not limited to
signaling motion vectors in the bitstream. Rather, the motion
vector may be determined by template matching already in the first
stage, correspondingly to the operation of the encoder. The
template matching of the first stage (motion vector derivation) may
be performed based on a search space different from the search
space of the second, refinement stage. In particular, the
refinement may be performed on a search space with higher
resolution (i.e. shorter distance between the search
positions).
[0094] An indication of the two reference pictures RefPic0 and
RefPic1, to which respective MV0 and MV1 point, are provided to the
decoder as well. The reference pictures are stored in the decoded
picture buffer at the encoder and decoder side as a result of
previous processing, i.e. respective encoding and decoding. One of
these reference pictures is selected for motion vector refinement
by search. A reference picture selection unit of the apparatus for
the determination of motion vectors is configured to select the
first reference picture to which MV0 points and the second
reference picture to which MV1 points. Following the selection, the
reference picture selection unit determines whether the first
reference picture or the second reference picture is used for
performing of motion vector refinement. For performing motion
vector refinement, the search region in the first reference picture
is defined around the candidate position to which motion vector MV0
points. The candidate search space positions within the search
region are analyzed to find a block most similar to a template
block by performing template matching within the search space and
determining a similarity metric such as the sum of absolute
differences (SAD). The positions of the search space denote the
positions on which the top left corner of the template is matched.
As already mentioned above, the top left corner is a mere
convention and any point of the search space such as the central
point can in general be used to denote the matching position.
[0095] According to the above mentioned document JVET-D0029, the
decoder-side motion vector refinement (DMVR) has as an input the
initial motion vectors MV0 and MV1 which point into two respective
reference pictures RefPict0 and RefPict1. These initial motion
vectors are used for determining the respective search spaces in
the RefPict0 and RefPict1. Moreover, using the motion vectors MV0
and MV1, a template is constructed based on the respective blocks
(of samples) A and B pointed to by MV0 and MV1 as follows:
Template=function((Block A,Block B)).
[0096] The function may be sample clipping operation in combination
with sample-wise weighted summation. The template is then used to
perform template matching in the search spaces determined based on
MV0 and MV1 in the respective reference pictures 0 and 1. The cost
function for determining the best template match in the respective
search spaces is SAD(Template, Block candA'), where block candA' is
the candidate coding block which is pointed by the candidate MV in
the search space spanned on a position given by the MV0. FIG. 3
illustrates the determination of the best matching block A' and the
resulting refined motion vector MV0'. Correspondingly, the same
template is used to find best matching block B' and the
corresponding motion vector MV1' which points to block B' as shown
in FIG. 3. In other words, after the template is constructed based
on the block A and B pointed to by the initial motion vectors MV0
and MV1, the refined motion vectors MV0' and MV1' are found via
search on RefPic0 and RefPic1 with the template.
[0097] Motion vector derivation techniques are sometimes also
referred to as frame rate up-conversion (FRUC). The initial motion
vectors MV0 and MV1 may generally be indicated in the bitstream to
ensure that encoder and decoder may use the same initial point for
motion vector refinement. Alternatively, the initial motion vectors
may be obtained by providing a list of initial candidates including
one or more initial candidates. For each of them a refined motion
vector is determined and at the end, the refined motion vector
minimizing the cost function is selected.
[0098] It is further noted that the present invention is not
limited to the template matching as described above with reference
to FIG. 3. FIG. 4 illustrates an alternative template matching
which is also applicable for uni-prediction. Details can be found
in document JVET-A1001, an in particular in Section "2.4.6. Pattern
matched motion vector derivation" of document JVET-A1001 which is
titled "Algorithm Description of Joint Exploration Test Model 1",
by Jianle Chen et. al. and which is accessible at:
http://phenix.it-sudparis.eu/jvet/. The template in this template
matching approach is determined as samples adjacent to the current
bock in the current frame. As shown in FIG. 1, the already
reconstructed samples adjacent to the top and left boundary of the
current block may be taken, referred to as "L-shaped template".
[0099] FIG. 5 illustrates another type of motion vector derivation
which may also be used. The input to the motion vector derivation
process is a flag that indicates whether or not the motion vector
derivation is applied. Implicitly, another input to the derivation
process is the motion vector of a neighboring (temporally or
spatially) previously coded/reconstructed block. The motion vectors
of a plurality of neighboring blocks are used as candidates for the
initial search step of motion vector derivation. The output of the
process is MV0' (possibly also MV1', if bi-prediction is used) and
the corresponding reference picture indices refPict0 and possibly
refPict1 respectively. The motion vector refinement stage then
includes the template matching as described above. After finding
the refined one (uni-prediction) or more (bi-prediction/multi-frame
prediction) motion vectors, the predictor of the current block is
constructed (for bi/multi-prediction by weighted sample prediction,
otherwise by referring to the samples pointed to by MV
refined).
[0100] The present invention is not limited to the 2 template
matching methods described above. As an example a third template
matching method which is called bilateral matching (also described
in the document JVET-A1001), can also be used for motion vector
refinement and the invention applies similarly. According to
bilateral matching, best match between two blocks along the motion
trajectory of the current block in two different reference pictures
is searched. Under the assumption of continuous motion trajectory,
the motion vectors MV0 and MV1 pointing to the two reference blocks
shall be proportional to the temporal distances, i.e., TD0 and TD1,
between the current picture and the two reference pictures. In
bilateral matching a cost function such as SAD(Block cand0', Block
cand1') might be used where Block cand0' is pointed by MV0 and
Block cand1' is pointed by MV1.
[0101] The processing circuitry 600 is illustrated in FIG. 6. The
processing circuitry may include any hardware and the configuration
may be implemented by any kind of programming or hardware design of
a combination of both. For instance, the processing circuitry may
be formed by a single processor such as general purpose processor
with the corresponding software implementing the above steps. On
the other hand, the processing circuitry may be implemented by a
specialized hardware such as an ASIC (Application-Specific
Integrated Circuit) or FPGA (Field-Programmable Gate Array) of a
DSP (Digital Signal Processor) or the like.
[0102] The processing circuitry may include one or more of the
above mentioned hardware components interconnected for performing
the above motion vector derivation. The processing circuitry 600
includes computation logic which implements two functionalities:
obtaining the initial motion vector (or a plurality of initial
motion vectors if bi-/multi-prediction is used) and template 610
and motion vector refinement 620. These two functionalities may be
implemented on the same piece of hardware or may be performed by
separate units of hardware such as initial motion vector and a
template determination unit 610 and motion vector refinement unit
620. The processing circuitry 600 may be communicatively connected
to an external memory 650 in which the reconstructed reference
picture samples are stored. Moreover, the processing circuitry 600
may further include an internal memory 640 which buffers the
samples in a window transferred from the external memory and used
for the motion vector determination for the currently processed
block. The processing circuitry may be embodied on a single chip as
an integrated circuit.
[0103] It is noted that the processing circuitry may implement
further functions of the encoder and/or decoder described with
reference to FIGS. 1 and 2. The internal memory may be an on-chip
memory such as a cache or a line memory. Chip memory is
advantageously implemented on the encoder/decoder chip to speed up
computations. Since the size of the chip is limited, the on-chip
memory is usually small. On the other hand, the external memory can
be very large in size, however the access to external memory
consumes more energy and the access is much slower. Usually the all
necessary information is retrieved from the external memory to
on-chip memory before the computations are performed. Worst case
external memory access (or bandwidth that needs to be provisioned
when designing the memory bus), denotes the largest possible amount
of memory transfer between external memory and the chip, while
decoding a frame or coding unit. The memory (especially the
external memory) can usually only be accessed in predefined block
units. In other words it is generally not possible to access a
single pixel, instead a smallest unit (e.g. 8.times.8) must be
accessed. The on-chip memory size is also an important design
consideration, as a larger on chip memory increases the cost.
[0104] In other words, the above mentioned apparatus may be an
integrated circuit further comprising: an internal memory embedded
within the integrated circuit and a memory access unit (interface)
for fetching integer samples located within said window from an
external memory to the internal memory.
[0105] The term "prediction block" employed above refers to the
current block which is to be predicted. It is a block within the
image which may be obtained by subdividing the image into equally
sized or differently sized (for instance by hierarchical
partitioning of a coding tree unit, CTU into the smaller units)
blocks. The block may be square or more generally rectangular as
these are the typical shapes also employed in current
encoders/decoders. However, the present disclosure is not limited
by any size/shape of the block.
[0106] The apparatus including the processing circuit may be the
encoder or decoder or even an apparatus including such encoder or
decoder, for instance a recording device and/or a playback
device.
[0107] Fractional sample positions are positions between the real
picture sample positions obtained by reconstructing the reference
picture which was coded as show in FIG. 1. Thus, the fractional
positions must be obtained by interpolation based on the nearest
integer positions. Details of interpolation filtering can be found
in Section "5.3 Fractional Sample Interpolation" of High Efficiency
Video Coding (HEVC) book by V. Sze et. al., Springer, 2014.
[0108] Interpolation filtering usually applies different filters in
order to generate different fractional pel (sample) positions. As
an example following 1D separable filters are applied to generate
quarter pel and half pel positions in H.265 video compression
standard:
TABLE-US-00001 Phase Luma filter coefficients 1/4 [-1, 4, -10, 58,
17, -5, 1]/64 1/2 [-1, 4, -11, 40, 40, -11, 4, -1]/64
[0109] As can be seen from the above table, the interpolation
filtering requires several samples around the fractional pel
position, corresponding to the filter taps (number of coefficients
in the table). Using the example filters above in order to generate
a half pel position 4 integer samples from left/top and
right/bottom are required. It should be noted that the length of
the interpolation filter is different for quarter pel sample
positions (which is 7 tap) than the half pel sample positions
(which is 8 tap). The terms sample, pel and pixel are used
interchangeably in this application, denoting an image sample (in
one or more color components) at a given time instance. A sample is
in principle an intensity value.
[0110] In some embodiments of the invention, the interpolation
filter of a predefined tap-size assesses integer samples only
within a window given by integer samples accessible for the
template matching in said search space. The window might include
much more samples than the ones actually used in computations of a
certain prediction block. This is due to the fact that the
refinement search operation is usually implemented using a fast
search method (as opposed to the brute force search method),
according to which some of the samples are not evaluated depending
on the gradual progression of the search operation. As a result the
number of template matching iterations as well as the samples that
are used in computations for refinement search operation might
change for each prediction block.
[0111] FIG. 7 illustrates a coding block (prediction block) and the
corresponding samples of the window. It is noted that the samples
shown in FIG. 7 are reference picture samples and the coding block
here is actually a block corresponding in size and position to the
current block in the current frame for which the motion vector is
to be derived in the reference picture. Thus, in fact, the coding
block in FIG. 7 is in fact a block co-located to the block for
which the predictor is searched. However, for the simplicity
reason, this block is referred as "coding block" in the
following.
[0112] In this example, unrefined motion vector MV0 points to an
integer sample position. The motion vector refinement search
granularity is 1 integer sample, meaning that since the starting
point is an integer sample, only integer sample points are
searched. The search is performed, in this example, in a gradually
developing search space. This means that the search space is in
each iteration advanced by adding new search positions depending on
the best direction in terms of cost function for the previously
tested positions.
[0113] Such approach is illustrated in a simplified manner in FIG.
8. In FIG. 8, the initial motion vector pointed to the center point
810. The search space is gradually constructed around the initial
motion vector position. In the first step, four positions
immediately adjacent on the top, bottom, left and right to the
position 810 pointed to by the initial motion vector as well as the
position 810 pointed to by the initial motion vector are tested.
Based on the direction which results in a lowest cost function
among the tested five points, further positions to be tested are
added to the search space. In this example, the lowest cost
function could be seen in the right point and so the search space
was extended by three further points in the horizontal right
direction in the second step. In the second step the lowest cost
function could be seen in right point (with respect to the lowest
cost point of the first step), resulting in a further extension of
the search space by three points in the horizontal right direction.
In the third step the lowest cost function is observed again in the
right point with respect to the lowest cost point of step 2 and
results in the extension of the search space by three more points
in the horizontal right direction. According to the example in FIG.
8, three more steps are performed in the top, top and right
directions in that order. In the example a diamond shaped pattern
(consisting of 5 search points) is used for each iteration and the
search space is extended in order to complete the missing search
points at each step.
[0114] In each iteration of the search space determination, the
search space may grow by one or more integer sample position.
Returning now to FIG. 7 in the example of which the maximum number
of search iterations is 4. Since the maximum number of 4 iterations
are possible, all of the samples depicted on the left need to be
retrieved from the memory to perform the search operation, in case
the gradual development of the search space goes to the left.
Similarly, 4 samples extension to the top is needed. Thus, the
search space is extended in both directions (left-right and
top-bottom) since the refined MV can move in either direction and
the hardware implementations require that all of the samples that
might be required are fetched from external memory before the
application of refinement search. If the search space develops in
the bottom or right direction, extension by 4 further samples is
necessary since the template matching with a template corresponding
to the size of the coding block (prediction block) will need to
access some of those samples. Moreover, the corner samples (e.g.
top-right) must also be fetched from the memory, since hardware
implementations typically cannot fetch irregular shapes
(rectangular access is more feasible).
[0115] It is noted that the above described iterative search space
development is only exemplary and the rules and number of points to
extend the search space in each iteration may differ, i.e. be
specified in a different way.
[0116] FIG. 8 also shows a scenario which may occur due to the
external memory access rules described above. The number of samples
that are fetched from the external memory is much higher than the
samples that are actually used in the computation step. Assuming
that the template here is only one sample large (for simplicity
reason), the white circles represent samples that are retrieved
from the external memory and the shaded samples that are actually
used. However, such redundancy is necessary if the number of
accesses to the external memory is to be kept low since when the
current block is started to be processed, the actually needed
samples are not yet known.
[0117] It is noted that the search space may also be defined in a
different way, for instance as a stabile shape located at the
position pointed to by the initial motion vector. The shape may be
any shape such as square, rectangle, diamond, or the like.
[0118] FIG. 9 illustrates an example in which the search space may
also include fractional samples. In FIGS. 7 and 8, the motion
vector search was performed on integer samples resulting in the
positions indicated by solid-line larger dots included in the
access window. If now the search is performed on a sample that has
half-pel resolution (smaller solid-line dot), in order to generate
the fractional sample, depicted on the left hand side, three more
columns of samples need to be retrieved from the memory as well,
assuming that the interpolation filter is symmetric and has eight
taps. Moreover same must be applied on the left side (extension by
3 columns of pixels) due to the fact that search operation is
symmetric (can move iteratively to the left and right) so that a
fractional pixel may be located on the left side of the window.
[0119] As a result due to interpolation filtering the number of
samples necessary to be retrieved from the memory is further
increased, indicated by the dashed line now also included the
dotted-line circles representing the positions added due to
fractional interpolation. Similarly if one allows half per
positions in the vertical direction to be searched as well, the
window of samples to be accessed from the memory needs to be
extended in the vertical direction too (not shown in the example of
FIG. 9), on the top and bottom sides.
[0120] Window of memory access is defined as the rectangle that
encloses all of the samples that need to be retrieved from the
memory in order to perform the motion vector search for a
prediction block (coding block). The window of memory access not
only includes the actual samples that are required, but also all of
the remaining samples that have the possibility of being accessed
during the motion vector search operation. In the example of FIG.
9, the motion vector search moved to the right. But it could have
been moved to the left direction as well, which is not known
beforehand. Accordingly, in order not to access the external memory
several times, the window of memory access (or access window)
includes all samples accessible by the respective processing.
[0121] FIG. 10 shows a fractional position 1010 pointed to by an
initial motion vector. In order to avoid search space construction
in which the points in integer distance from the initial motion
vector position would all be fractional, requiring interpolation of
all these points and thus further integer positions from the
memory, the fractional point 1010 is shifted to a nearest integer
point (shown as shaded circle in FIG. 10). In FIG. 10, the
fractional position 1010 to which the initial motion vector points
is a half-pel position between two integer points, with respect to
which it is located in the same distance. In such situation there
can be a convention defined, which of the two equally distanced
integer points are to be used. In this example, the rule left
before right was followed. Similarly, a rule top before bottom may
be used and horizontal before vertical, resulting in the preferred
sequence: left, top, right, bottom. In other words, if the
fractional position has the same distance to 4 integer positions,
it will be rounded to the left top position. However, these are
only examples. In general, the order may be any other order as long
as it is identified so that the encoder and decoder may handle
accordingly.
[0122] If there are more than one integer samples that are at the
same distance to the initial motion vector, another rule that can
be applied might be to choose the sample coordinate that is closest
to the (0,0) center coordinate. This rule makes sense since the
probability of observing a motion vector with a bigger magnitude
(in either direction, along the x- and y-axes) is statistically
less than a motion vector with a smaller magnitude. In other words
video sequences usually contain less motion and still backgrounds
(that can be predicted with smaller magnitude motion vectors), than
chaotic motion and moving backgrounds (that can be predicted with
bigger magnitude motion vectors).
[0123] According to the rounding rule that is presented in the
above paragraph, is the vertical component (along the y-axis) of an
initial motion vector is 5.5 and the horizontal component is -3.5,
it is rounded as (5,3). According to the rounding rule, if the
initial motion vector is closest to more than one integer sample
coordinates, the initial motion vector is rounded to achieve the
resulting rounded motion vector with minimum absolute magnitude
along both x- and y-axes. Such rounding may be, for instance,
implemented by simply cutting of the positions after comma, i.e. by
function "floor".
[0124] In other words, if there are more than one integer sample
positions equally close to the fractional position, the rounding is
obtained as the one among the equally close integer sample
positions which results in a shorter magnitude of the rounded
initial motion vector.
[0125] In one embodiment, the search space consists of sample
positions spaced by integer sample step from each other, as also
shown in FIG. 10, in which the search space window 1000 frames the
integer samples of the search space constructed around the center
position (shaded circle) corresponding to the rounded and thus
integer position after shifting the fractional initial motion
vector position as described above.
[0126] In order to ensure that refinement of the motion vector does
not worsen the initial motion vector, in one exemplary embodiment,
the template matching cost for the initial motion vector pointing
to the fractional sample position is calculated and compared with
the template matching cost of the determined refinement of the
initial motion vector. Finally, the motion vector for the
prediction block is determined based on the fractional sample
position if the template matching cost of the determined refinement
is not lower than the calculated template matching cost of the
fractional sample position. In other words, the refinement is only
selected for the prediction block if it is better (in terms of the
cost function) than the initial motion vector which means that the
motion vector for the prediction block is determined based on the
determined refinement if the template matching cost of the
determined refinement is lower (or equal to) than the calculated
template matching cost of the fractional sample position.
[0127] On one hand, the motion vector for the prediction block may
be determined to be directly the fractional sample position if the
template matching cost of the determined refinement is not lower
than the calculated template matching cost of the fractional sample
position. However, there might be further steps (e.g. a following
template search in the fractional search space, motion vector
clipping or the like) that further modifies the motion vector.
[0128] The search space may be formed by K rows and L columns of
samples located in rows and columns in integer distance from each
other, wherein the rounded fractional position is one of the
samples of the search space. Here, K and L are integers larger than
1. As shown in FIG. 10, K and L may have equal value (9 in this
example). However, in general, K and L may differ in value, i.e.
the search space may be rectangular with different side
lengths.
[0129] The term search space generally refers to a set of positions
on each of which template matching is to be performed. These are
the positions that are actually checked by the template matching,
i.e. for each search space position, one respective cost function
value is calculated. In contrast, the memory access window (also
referred to as access window or window in this disclosure) includes
the search space but may also include samples which are to be
accessible for template matching, i.e. may be used to calculate
cost function for at least one of the above mentioned search space
positions. In the above example of search space K.times.L, the
memory access window may have the same or larger size.
[0130] The rounded fractional position may be located substantially
in the center of the search space or of the memory access window or
both, as is also shown in FIG. 10. The term "substantially" here
refers to the fact that if the number of rows and/or columns of the
search space and/or memory access window is even, there is no
integer center. In such case, the rounded fractional position would
be located on one of the positions closest to the center.
[0131] However, it is noted that the present disclosure is not to
be limited to the above mentioned search space configurations. In
general, any search space (for instance an iteratively obtained
search space as described above) may be employed. Moreover, the
initial motion vector is not necessarily in the center of the
search space. For instance, this is not the case for the iterative
search space in which the search space develops in several
iterations in direction of the lowest cost function thus extending
in one particular direction from the initial motion vector, which
is then not in the center thereof.
[0132] The window for memory access may be defined in various ways.
FIG. 11 illustrates an example in which the memory access window is
defined as extensions EXT on the left/right or up/down of the
coding block (corresponding to the location of the coding block
given by the initial motion vector). The extension amounts may
depend on the size and the shape of the coding or prediction block.
In FIG. 11, the extension is 4 samples long in each direction (top,
bottom, left, right). However, it is noted that the EXT may also
take different values for the different directions, depending on
the block size (which may have different size in vertical and
horizontal direction) and/or the search space form and size.
[0133] For instance, according to an example, the window is defined
as N integer sample columns and M integer sample rows relative to
the prediction block initial motion vector, N and M being non-zero
integer values. In FIG. 11, the N and M are indicated but have the
same size. As mentioned above, N and M may have a different size. N
and M are integers and at least one of them are non-zero. Taking
parameters N and M and the template form and size, the size of the
access window can be determined. In particular, if the template has
T1 rows and T2 columns, the size of the memory access window may be
calculated as (N+T1+N) rows and (M+T2+M) columns. This is because
the search can go N samples left or right resulting in 2N samples
horizontally and M samples up or down resulting in 2M samples
vertically. In addition, the template can be matched with
T1.times.T2 samples on any of the 2N.times.2M sample positions.
[0134] On the other hand, for specific approaches of search space
constructions as the one described with reference to FIGS. 7 and 8,
the memory access window can be defined in terms of maximum number
of refinement iterations (search space construction iterations) and
iteration step size (in terms of maximum sample distance achievable
in each iteration), which can be later converted to maximum amount
of displacement in left, right, up and down. Therefore the memory
access window is defined as the maximum displacement in each
direction. For instance, the 4 iterations in which each iteration
may advance the search space in maximum one integer sample position
result in EXT=4.
[0135] In other words, according to this example, the processing
circuitry is configured to determine the refinement of the initial
motion vector by template matching with said template in a search
space which is iteratively extended in a direction given by one (or
more) of more best matching positions of the search space in a most
recent iteration, the window is defined by a predefined maximum
number of the iterations.
[0136] It is noted that in general, the present disclosure is not
limited to any particular shape or form or kind of determination of
the search space. In another example, search space is a rectangular
sub-window of the window or may be equal to the window. The
definition of the memory access window may be relevant in order to
possibly include the corresponding signaling parameter into the
bitstream (for instance parameter EXT or parameters N and M).
However, the memory access window size may also be defined in the
standard or derivable on the basis of other coding parameters (such
as template size, prediction block size, image resolution, etc.).
As already noted above, the template matching is not necessarily
performed in a rectangular search space including all positions. In
an alternative embodiment, the template matching performed for a
search space of which the sample positions are determined in a
plurality of steps (iteratively) including the steps of:
determining an integer search space consisting of sample positions
spaced by integer sample step from each other, performing template
matching in the integer search space to obtain a best matching
integer position, based on the best matching integer position,
determining a fractional search space with at least one sample
spaced by less than integer sample step from the closest integer
search space sample positions, and performing template matching in
the fractional search space to obtain a best matching position. In
other words, the search space which may be in the beginning and
possibly several iterations limited to samples in integer distance
from each other in rows and columns, may further develop by adding
some fractional positions.
[0137] A search space construction which is performed iteratively
according to the cost function and also includes fractional
position enables reduction of the number of template matching
operations to be performed to find the motion vector refinement.
Moreover, it may still increase the accuracy by providing also
fractional positions to be checked.
[0138] For example, the rounding may be obtained as an integer
sample position closest to the fractional position to which the
initial motion vector points. If there are more than one integer
sample position equally close to the fractional position, the
rounding is obtained as an integer sample position in a predefined
direction, the predefined direction or combination of directions,
which may be left, right, top or bottom. As explained above, a
fractional position in an integer sample search space can have
either two or four closest neighbors. The predefined direction may
be specified by ordering the four possible directions in a
preferred order so that among two directions of the two equally
distanced respective integer samples the one is selected which is
listed as first of them in the predefined ordering. A combination
of directions is relevant, if the fractional position is not on
line with the integer samples but rather shifter in horizontal as
well as vertical direction. In such case, the fractional position
has to be rounded to an integer position in both directions such as
top left or top right or the like.
[0139] However, the present disclosure is not limited to a single
rounding possibility. Alternatively, if there are more than one
integer sample positions equally close to the fractional position,
the search space is a unity of a search space located on said more
than one integer sample positions equally close to the fractional
position. In other words, the search space is a combination of
search spaces obtained by respective shifting to all (two or four)
equally close integer sample positions. In practice this will
result, for instance in one additional column and/or row in the
resulting search space as compared to the case where only single
shift is performed for two equally spaced integer positions.
[0140] Another alternative rounding rule that can be applied might
be to choose the sample coordinate that is closest to the (0,0)
center coordinate. The center coordinate is in general the
coordinate from which the initial motion vector starts. Typically,
it would be the position of the collocated block, i.e. a coding
block in the reference picture corresponding to the position of the
prediction block for which the predictor is to be determined in the
current frame. Accordingly if the initial motion vector is equally
close to more than one integer sample coordinates, the initial
motion vector is rounded to achieve a resulting rounded motion
vector with minimum absolute magnitude along both x- and
y-axes.
[0141] FIG. 12 shows on the left hand side a typical approach in
which the initial motion vector position points to a position
(starting coordinate of the motion vector) 1210 which is
fractional. The motion vector refinement process is usually applied
using fixed step sizes (distance difference between search points).
Typically, the distance is integer sample distance, i.e.
corresponding to a distance between two integer samples. This means
that if the starting point of the search operation is a fractional
pel-point, then consecutive search points are also fractional-pel
points which is illustrated in FIG. 12 by several fractional
positions (smaller bold circles). In the example of FIG. 12, not
all fractional positions are shown, in principle, if the search
space is defined as a rectangle around the initial motion vector,
there would be a fractional position corresponding to each of the
grid of the integer samples shown in FIG. 12. As can be seen from
the figure, interpolation of the fractional positions may require
values of sample positions 1220 out of the search space. For
instance, if 6-tap filter is used, two more columns on the right
side of the search space would be necessary. If the memory access
window is considers to be the same as search space as in this
example, this would require additional memory access to fetch the
additional samples or a larger storage to internally store two more
columns of samples. It is noted that in general, the search space
and the memory access window are not the same. FIG. 12 is a
simplified example, in which the template is assumed to have size
1.times.1. In general, the search space would be smaller than the
memory access window. The memory access window would additionally
include positions which do not belong to search space positions but
are used for template matching. The number of additional pixels
would depend on the template size. Moreover, the fractional
interpolation would only need additional sample positions compared
to the template matching, if the interpolation filter size divided
by two was larger than template size. However, it is noted that the
rounding of the initial motion vector does not only provide the
advantage of reducing the window size in such cases. In addition,
it also reduces the number of operations which are necessary to
interpolate the fractional positions to obtain the search
space.
[0142] On the right hand side in FIG. 12, the shifting (rounding)
of the fractional position 1210 to an integer position is shown. As
a result, the search space constructed based on the integer
position corresponding to the rounded fractional position includes
integer samples and does not require any additional fractional
position interpolation in this example.
[0143] In other words, in the state of the art since the starting
point is a half pel position and since the step size of refinement
search step is typically 1 pixel (sample), always fractional
positions are searched. As a result, since interpolation filtering
requires samples outside of the search space/memory window from the
right, the worst case memory access is high. According to an
embodiment of the invention first the starting coordinate is
rounded to an integer starting point. As a result no sample outside
of the memory access window need to be accessed even if the search
coordinate is at the border (still inside) of the memory access
window.
[0144] The main benefit of the fractional initial motion position
is that given a predefined memory access window the rounding helps
the search operation reach a farther point than the initial
starting point (since the integer search coordinates do not require
interpolation filtering hence additional samples. According to a
different viewpoint, if the same number of refinement iterations
are applied, the memory access requirement of the invention is
smaller than the state-of-the-art.
[0145] It is noted that the above examples were provided for
half-pel interpolation. However, the present disclosure is not
limited thereto. In general, any fractional position such as 1/4,
1/8, or the like may be used, i.e. interpolated using the
corresponding interpolation filter.
[0146] The processing circuitry described with reference to FIG. 6
may be employed in an encoder and/or decoder as shown in FIGS. 1
and 2.
[0147] In particular, an encoding apparatus may be provided for
encoding video images split to prediction blocks into a bitstream,
the encoding apparatus comprising: the apparatus for determination
of a motion vector for a prediction block as described above
including the processing circuitry; and an encoding circuitry for
encoding difference between the prediction block and the predictor
given by a prediction block in a position specified by the
determined motion vector and for generating bitstream including the
encoded difference and the initial motion vector.
[0148] Further units and functions of the encoder described above
with reference to FIG. 1 may also be provided or implemented in the
processing circuitry.
[0149] Correspondingly, a decoding apparatus may be provided for
decoding from a bitstream video images split to prediction blocks,
the decoding apparatus comprising: a parsing unit for parsing from
the bitstream an initial motion vector and an encoded difference
between a prediction block and a predictor given by a prediction
block in a position specified by a refined motion vector; the
apparatus for determination of the refined motion vector for the
prediction block as described above including the processing
circuitry; and a decoding circuitry for reconstructing the
prediction block as a sum of the parsed difference and the
predictor given by the prediction block in the position based on by
the refined motion vector. For example, the predictor may be
directly given by the position of the refined motion vector.
However, there may be further processing steps of obtaining the
motion vector of the current prediction block which may further
change the motion vector (such as filtering, clipping, further
refinement or the like).
[0150] Further units and functions of the decoder described above
with reference to FIG. 2 may also be provided or implemented in the
processing circuitry.
[0151] Moreover, the embodiments of the invention were described
from the point of view of the apparatus with the processing
circuitry to perform the motion vector refinement. However, the
present disclosure is not limited thereto but also provides the
corresponding methods which include the processing steps
corresponding to those for the performing of which the above
described processing circuitry is configured.
[0152] The taps corresponds to the filter coefficients. The
tap-size corresponds to filter order. Here, it is assumed that the
filter is a linear filter. In some examples, the filter may be
symmetric, i.e. having symmetric coefficients. However, the present
disclosure is not limited to symmetric filters or linear filters or
any kind of filters. In general, the fractional positions may be
obtained in any way based on the adjacent samples.
[0153] FIG. 13A illustrates a typical approach in which the initial
motion vector is input to the motion vector refinement. Then a
search space is determined with positions located in integer sample
distance from the initial motion vector and the search is performed
therein. This results in a search space with only fractional
positions which are all to be interpolated in case the initial
motion vector has a fractional coordinate. Additionally, some
fractional-distance samples may be added to the search space and
also tested by the template matching to find the best match for the
prediction block.
[0154] FIG. 13B illustrates an approach according to the present
disclosure. In particular, an initial (non-refined) motion vector
is input to the motion vector refinement procedure. If the position
to which the initial motion vector points (i.e. at least one the
motion vector coordinate) is fractional, then the position (the
fractional one or both coordinate(s)) is rounded to the next
closest integer position. The search space is then determined as in
case or FIG. 13A. However, due to the rounding, the search space is
forms by integer positions.
[0155] After the integer-distance search a finer,
fractional-distance search may be performed around the motion
vector refinement in the integer-distance search space. The present
disclosure does not disallow a later motion vector refinement
search based on fractional search coordinates, given that the
defined memory access window is not exceeded (due to additional
interpolation operation). Usually the refinement operation is
applied in progressively smaller step sizes, once the integer step
size search is complete, finer step sizes might also be checked. As
a special case, the Initial Non-Refined MV might always be checked
and compared with the remaining search coordinates, since it is a
candidate coordinate that was selected as the best motion vector by
a previous operation.
[0156] FIG. 13C illustrates an alternative approach in which the
rounding is performed in different ways resulting in different
overlapping search spaces. In particular, More than one rounding
alternative can be generated by the motion vector rounding step,
especially if there are more than one closest integer sample
position. Accordingly in order to reduce the memory access the
starting search point is rounded to an integer-pel point (more
generally to a position whose reconstruction requires less pixel
extension) or more integer-pel points (typically 2 or 4). Therefore
the subsequent search points are also integer pel points.
[0157] In the above embodiments and examples, by shifting a
fractional motion vector position to an integer position, a search
space may be constructed which mainly includes integer samples.
However, still, such shifted search space based on the rounded
initial motion vector rather than fractional motion vector, may
include some fractional positions. For example, if the search space
is determined iteratively as illustrated with reference to FIG. 8,
then in a last iteration, one or more fractional positions may be
tested. However, this is only an example and the fractional
positions may be tested in any iteration. In order to avoid that
fractional positions require a repeated access to the memory to
fetch additional samples which may be necessary for fractional
interpolation, in the following, two further embodiments are
described, namely: [0158] Limiting the memory access window and
only enabling fractional positions in the search space template
matching operation, which do not require extension of the window,
wherein the memory access window is defined in order to enable
accessing any samples necessary for template matching for a certain
prediction block [0159] Padding samples which extend over a memory
access window with some predetermined replacement values depending
on the samples within the window.
[0160] These embodiments will be described in detail in the
following. In particular, the processing circuitry may be further
configured accordingly, in order to maintain the memory access
window and number of its accesses low. The corresponding methods
and apparatuses are provided.
Limited Memory Access Window for Interpolation Filtering
[0161] As mentioned above, the external memory access is one of the
most important design considerations in today's hardware and
software architectures. Motion vector estimation especially when
including template matching, for instance in case of motion vector
refinement, may also be used with interpolation filtering to obtain
fractional positions of the search space. Use of the interpolation
filtering may require increase of the number of samples which need
to be accessed from the memory. However, this may lead to either
increase of the expensive on-chip memory or to increased number of
external memory accesses, which on the other hand slows down the
implementation. Especially at the decoder side, these problems may
lead to more expensive or slower applications which is not
desirable.
[0162] In order to prevent such situation, the present disclosure
in another embodiment provides a restriction on external memory
accesses. According to an embodiment of the invention, a window of
samples which are to be accessible for the motion vector refinement
is defined around the position pointed to by a non-refined motion
vector, i.e. by the initial motion vector. The window defines the
maximum number of samples that need to be accessed from the memory
in order to perform the motion vector refinement. In general, the
samples which are to be accessible include the samples on positions
of the search space in which the template matching is to be
performed and the samples which are to be matched with the template
for all positions in the search space. The latter typically exceed
the search space. For simplicity reasons the window of memory
access may be defined as an extension around the coding block
(current block for which the motion vector is to be found). For
example, R samples from left and right and R samples from top and
bottom boundary of the current block may define the window. In
other words, when the current block has a size of N.times.N
samples, the access window may have size of (R+N+R).times.(R+N+R),
i.e. (N+2R).times.(N+2R) samples For instance, R may be equal to 4.
However, the current block may be of vertical size N different from
the horizontal size N and the number of extension samples in top,
bottom, left and right direction may also differ.
[0163] According to the present disclosure, in order to limit the
memory access window, the fractional pixel coordinates are accessed
by the motion vector refinement only if samples necessary for the
interpolation lie inside the window of memory access for the motion
vector refinement as defined for integer samples.
[0164] FIG. 14 shows a window of memory access for motion vector
refinement. Center point 1410 is the position pointed to by the
non-refined input motion vector (initial motion vector obtained
either from the bitstream or by previously performed template
matching or testing of the candidates as described above). In order
to avoid further increasing of the window size due to adding
fractional positions to the search space, motion vector refinement
is performed according to following rules: [0165] A) A window of
memory access for refinement is defined around the non-refined
initial motion vector coordinate (i.e. position pointed to by the
initial motion vector). The window identifies the maximum number of
pixel samples that need to be accessed from the memory in order to
perform motion vector refinement by template matching in a search
space. [0166] 1. In this example, for the sake of simplicity, the
size of the current block (coding block size) is 1.times.1 sample,
but it can be larger and it typically would be larger. [0167] 2.
The window of memory access is defined as the extension around the
coding block, such as 4 samples from left/right and 4 samples from
top/bottom shown in the figure. [0168] B) The fractional pixel
coordinates are accessed for MV refinement only if samples
necessary for interpolation lie inside the window of memory
access.
[0169] Requirement B ensures that the access window defined by the
samples necessary for motion vector refinement on integer samples
is not further extended. The actual fractional samples accessible
according to this rule are given by the size and shape of the
interpolation filter. Accordingly, in FIG. 14, assuming an
interpolation filter with 6 taps, the dotted line indicates a
region in which the fractional samples may be located. However, it
is noted that further fractional pixel positions may be allowable
as will be shown with reference to FIG. 15. In particular,
fractional positions which require only vertical or only horizontal
filtering which does not require extension beyond the access window
may still be used. Thus, limiting the fractional positions to the
fractional sample window shown in FIG. 14 may be too limiting for
some applications.
[0170] In other words, according to an embodiment, the memory
access window includes all samples that are accessible by the
motion vector refinement performed on integer samples and do not
include samples that are not accessible by motion vector refinement
performed on integer samples. Thus, if fractional samples are used
for motion vector refinement, they are obtained in a manner which
does not require additional samples.
[0171] In the example of FIG. 14, this is achieved by only allowing
fractional samples which, for a predefined interpolation filter
shape and size, do not require samples out of the access window.
The dotted fractional sample window extends within the access
window. If T is the number of interpolation filter taps, then the
fractional sample window border is defined by integer samples in a
distance 1420 of floor(T/2)-1 from the access window border
samples. In particular, in this example T=6, T/2=3 and the distance
from the access window border sample to the fractional window
sample is T/2-1=2 integer samples.
[0172] However, it is noted that this determination of the
fractional window is a mere example. In general, the window may
have a different form and size. The vertical and horizontal
interpolation may be done by filters with different sizes.
Moreover, some fractional positions may require filter in both
vertical and horizontal direction which may in general by separable
or non-separable.
[0173] Alternatively, the interpolation filter could be changed
(e.g. number of taps at least in one direction reduced) for the
fractional positions outside the dotted window in FIG. 14. However,
for implementation purposes and for interpolation quality reasons,
such solution may be less attractive.
[0174] It is noted that in general, the present disclosure is not
limited to any particular shape or form or kind of determination of
the search space such as the iterative approach. In another
example, search space is a rectangular sub-window of the window
such that all integer samples accessed for interpolation filtering
of each fractional sample in the sub-window are located within said
window for the interpolation filter with the predefined tap-size. A
similar example has been already discussed above with reference to
FIG. 14. In FIG. 14, the search space is given by the 9.times.9
integer samples and fractional samples located within the region
formed by 5.times.5 integer samples with the initial motion vector
position in their center.
[0175] FIG. 15 illustrates an example with different fractional
pixel positions. For this example, it is assumed that the size of
the template is 1.times.1 samples (for simplicity) and 6 tap
interpolation filter is used for each half-pel position. In the
example, the search coordinates that are searched are denoted by
numbers 1-6 which indicate the order they are checked, i.e. the
order in which the template matching search is performed. Positions
1 and 2 are half-pel position (meaning that they are located in the
middle between two integer sample positions, pel is an abbreviation
for pixel and the term pixel is used interchangeably with the term
sample in this application). Positions 1 and 2 are checked since
the necessary extension for interpolation filtering lies inside the
window of memory access (3 integer samples on the diagonal top-left
and 3 integer samples bottom-right from position 1; 3 integer
samples to the right and 3 integer samples to the left of position
2). Note that fractional sample point 1 requires extension both in
horizontal and vertical direction, both of which lie inside the
window. Position 2 only requires extension to the right and to the
left.
[0176] Positions 3, 4 and 5 are integer sample (integer-pel)
positions. They can be searched since no extension for
interpolation filtering is necessary. Fractional (half-pel) sample
6 can also be accessed since only an extension in the vertical
direction (by three integer positions up and down respectively) is
necessary which is still inside the window. No extension in the
horizontal direction is necessary. Still, in the above
implementation fractional sample points are accessed only if
necessary interpolation extension is within the memory access
window.
[0177] In other words, according to an example, the interpolation
filter is a one-dimensional filter assessing K either horizontal or
vertical integer samples when the fractional position is located on
a respective horizontal or vertical line of integer samples.
[0178] Such one-dimensional fractional positions (e.g. positions 2
and 6 in FIG. 15) located on a line between two horizontally or two
vertically adjacent integer positions require an extension for
interpolation only in horizontal direction or only in vertical
direction, i.e. are to be filtered only by a respective horizontal
or vertical interpolation filter. In order to be able to make use
of as many fractional positions as possible, in addition to the
fractional positions allowed in the example of FIG. 14, it may be
advantageous to add further one-dimensional positions such as
position 6 shown in FIG. 15.
[0179] In other words, the search space further includes fractional
positions located outside the fractional sub-window (cf. dotted
window of FIG. 14) either: [0180] adjacent on the top or on the
bottom of the sub-window and located on the horizontal line of
integer samples or [0181] adjacent on the left or on the right hand
side of the sub-window and located on the vertical line of integer
samples.
[0182] It is noted that some fractional samples might require more
integer samples in a given direction, horizontal or vertical. This
may be the case if the predefined filter size is different to
generate that position in the respective directions.
[0183] FIG. 16 illustrates an example of fractional half-pel
positions 1 and 2 which cannot be accessed. They are located
outside the sub-window shown in FIG. 14. For this example, it is
assumed that 6 tap interpolation filter is used for half pel
positions. The half-pel search points 1 and 2 are not allowed to be
searched, since horizontal or vertical interpolation filtering
requires samples that lie outside of the window. The integer sample
positions needed by the horizontal filter to filter position 1 and
the vertical filter to filter position 2 are indicated by a dashed
line in FIG. 16. As can be seen, dashed circles correspond to the
integer positions which are not within the memory access
window.
[0184] In the above examples, the memory access window was defined
so that no sample outside of the window is accessed (even for
interpolation filtering) during the motion vector refinement
process. In other words memory access window is the smallest window
that encloses the samples that might need to be accessed for motion
vector refinement and interpolation. Moreover, the memory access
window has been designed according to the samples necessary for
motion vector refinement based on integer sample positions. Then,
only fractional positions are allowed which do not require further
extension of such access window by further integer positions. It is
noted that the above examples were provided for half-pel
interpolation. However, the present disclosure is not limited
thereto. In general, any fractional position such as 1/4, 1/8, or
the like may be used, i.e. interpolated using the corresponding
interpolation filter.
[0185] Moreover, method is provided for determination of a motion
vector for a prediction block including the steps of: obtaining an
initial motion vector and a template for the prediction block;
determining a refinement of the initial motion vector by template
matching with said template in a search space, wherein said search
space is located on a position given by the initial motion vector
and includes one or more fractional sample positions, wherein each
of the fractional sample positions belonging to the search space is
obtained by interpolation filtering with a filter of a predefined
tap-size assessing integer samples only within a window, said
window being formed by integer samples accessible for the template
matching in said search space. The taps corresponds to the filter
coefficients. The tap-size corresponds to filter order. Here, it is
assumed that the filter is a linear filter. In some examples, the
filter may be symmetric, i.e. having symmetric coefficients.
However, the present disclosure is not limited to symmetric filters
or linear filters or any kind of filters. In general, the
fractional positions may be obtained in any way based on the
adjacent samples.
[0186] FIG. 17 shows an exemplary implementation of a method
according to an embodiment. A function InterpolationFilterLength(C)
returns the number of additional samples necessary in the
horizontal and vertical directions in order to apply interpolation
filtering. The number of necessary samples changes depending on:
[0187] Whether the search coordinate is integer pel, half pel, or
quarter pel position. [0188] Whether horizontal, vertical or both
of the interpolation filters need to be applied to generate the
search coordinate sample.
[0189] The method starts in step 1730. In particular, the initial
motion vector position is the first search space position C(x,y) to
be tested. The function InterpolationFilterLength(C) returns for
this position the number of samples in the horizontal and vertical
directions in order to apply interpolation filtering. If the sum of
C(x,y) and InterpolationFilterLength(C(x,y)) exceeds the access
window size defined by MAX (max_x, max_y), then the position is not
used as a part of the search space. Instead, next search coordinate
C(x, y) is selected in step 1740 to be tested (for instance, x or y
or both are incremented, depending on the order in which the search
is performed). If the tested position in step 1730 does not require
exceeding the access window, the in step 1710 the template matching
as a part of the motion vector refinement is performed for that
position C(x,y). Then it is tested in step 1720, whether there are
still search space positions left for template matching. If not,
the refinement is terminated. If yes, the next coordinate is
selected in the step 1740 and the condition of step 1730 is
evaluated for that new position. These steps are repeated.
[0190] As already described above, alternative possibilities exist
to define the allowed fractional positions (such as the window in
FIG. 14, possibly extended by further fractional samples as shown
in FIG. 15). Based thereon, a simpler condition may be formulated
in step 1730 merely evaluating whether the position C(x,y) belongs
to the allowed window.
[0191] In other words, this embodiment is based on observation that
motion vector refinement when implemented in combination with
fractional interpolation may require further increase of on-chip
memory size or even external memory access. Both options may be
undesirable.
[0192] In view of the above mentioned problem, the embodiment
provides motion vector prediction which enables to take into
account the number of accesses to the external memory and the
number of samples which are necessary to be accessible for motion
vector refinement of a motion vector for a coding block.
[0193] This is achieved by limiting the number of samples to those
necessary for the integer sample template matching and only
enabling those fractional positions which are obtainable with a
predetermined interpolation filter without requiring additional
integer samples.
[0194] According to an aspect of the invention, an apparatus is
provided for determination of a motion vector for a prediction
block including a processing circuitry configured to: obtain an
initial motion vector and a template for the prediction block; and
determine a refinement of the initial motion vector by template
matching with said template in a search space. Said search space is
located on a position given by the initial motion vector and
includes one or more fractional sample positions, wherein each of
the fractional sample positions belonging to the search space is
obtained by interpolation filtering with a filter of a predefined
tap-size assessing integer samples only within a window, said
window being formed by integer samples accessible for the template
matching in said search space.
One of the advantages of such motion vector determination is
limited number of samples which need to be accessible for
performing the motion vector refinement for a prediction block
while at the same time limiting the number of accesses to the
external memory or generally to a memory/storage/cache storing the
entire reference pictures.
[0195] In an example, the window is defined as N integer sample
columns and M integer sample rows relative to the prediction block
initial motion vector, N and M being non-zero integer values. Such
definition may provide a simple means for specifying which samples
are to be retrieved for the purpose of motion vector determination
and/or refinement. It may also be easily configurable for instance
within a bitstream or standard.
[0196] In one embodiment, the processing circuitry is configured to
determine the refinement of the initial motion vector by template
matching with said template in a search space which is iteratively
extended in a direction given by one of more best matching
positions of the search space in a most recent iteration, and the
window is defined by a predefined maximum number of the
iterations.
[0197] The search space may include a rectangular sub-window of the
window such that all integer samples accessed for interpolation
filtering of each fractional sample in the sub-window are located
within said window for the interpolation filter with the predefined
tap-size.
[0198] In particular, as a specific example, the interpolation
filter is a one-dimensional filter assessing K either horizontal or
vertical integer samples when the fractional position is located on
a respective horizontal or vertical line of integer samples.
[0199] Moreover, for instance, the search space further includes
fractional positions located outside the sub-window either: [0200]
adjacent on the top or on the bottom of the sub-window and located
on the horizontal line of integer samples or [0201] adjacent on the
left or on the right hand side of the sub-window and located on the
vertical line of integer samples.
[0202] According to another aspect of the invention, an encoding
apparatus is provided for encoding video images split to prediction
blocks into a bitstream, the encoding apparatus comprising: the
apparatus for determination of a motion vector for a prediction
block as described above; and an encoding circuitry for encoding
difference between the prediction block and the predictor given by
a prediction block in a position based on the determined motion
vector and for generating bitstream including the encoded
difference and the initial motion vector.
[0203] According to another aspect of the invention a decoding
apparatus is provided for decoding from a bitstream video images
split to prediction blocks, the decoding apparatus comprising: a
parsing unit for parsing from the bitstream an initial motion
vector and an encoded difference between a prediction block and a
predictor given by a prediction block in a position specified by a
refined motion vector; the apparatus for determination of the
refined motion vector for the prediction block as described above;
as well as a decoding circuitry for reconstructing the prediction
block as a function of the parsed difference and the predictor
given by the prediction block in the position specified by the
refined motion vector. The function may be or include a sum. The
function may further comprise clipping, rounding, scaling or
further operations.
[0204] According to another aspect of the invention a method is
provided for determination of a motion vector for a prediction
block including the steps of: obtaining an initial motion vector
and a template for the prediction block; determining a refinement
of the initial motion vector by template matching with said
template in a search space, wherein said search space is located on
a position given by the initial motion vector and includes one or
more fractional sample positions, wherein each of the fractional
sample positions belonging to the search space is obtained by
interpolation filtering with a filter of a predefined tap-size
assessing integer samples only within a window, said window being
formed by integer samples accessible for the template matching in
said search space.
[0205] For instance, the window is defined a as N integer sample
columns and M integer sample rows relative to the prediction block
initial motion vector, N and M being non-zero integer values.
[0206] In an embodiment, the refinement of the initial motion
vector is determined by template matching with said template in a
search space which is iteratively extended in a direction given by
one of more best matching positions of the search space in a most
recent iteration, the window is defined by a predefined maximum
number of the iterations.
[0207] In an exemplary implementation, the search space includes a
rectangular sub-window of the window such that all integer samples
accessed for interpolation filtering of each fractional sample in
the sub-window are located within said window for the interpolation
filter with the predefined tap-size.
[0208] Moreover, for instance, the interpolation filter is a
one-dimensional filter assessing K either horizontal or vertical
integer samples when the fractional position is located on a
respective horizontal or vertical line of integer samples.
[0209] Advantageously, the search space further includes fractional
positions located outside the sub-window either: adjacent on the
top or on the bottom of the sub-window and located on the
horizontal line of integer samples or adjacent on the left or on
the right hand side of the sub-window and located on the vertical
line of integer samples.
Padding
[0210] Alternatively to the limiting of the fractional positions in
the search space based on the memory access window, the present
disclosure performs in its another embodiment padding of the
samples which are located outside the memory access window if these
are needed, for instance, to interpolate fractional samples.
[0211] The provision of replacement samples (padding samples) may
be further advantageous in connection with template matching on
search spaces which also include fractional samples. In other
words, the search space for a prediction block may include at least
one fractional sample position with a value obtained by
interpolation filtering with a filter of a predefined tap-size. The
interpolation filtering uses sample position values from the window
and the at least one replacement value to obtain the at least one
fractional sample position value.
[0212] FIG. 18 shows a window of memory access for the purpose of
motion vector refinement The window is spanned around the point
1810 in the center which is the position pointed to by the initial
motion vector. Moreover, fractional position 1820 also belongs to
the search space. However, the fractional position 1820 is to be
interpolated by a 6 tap filter using 3 integer sample positions to
the left of the fractional position and three integer sample
positions to the right of the fractional position. Since the
fractional position 1820 is located on a line with the integer
sample positions, a one-dimensional filter may be applied for
interpolation, for instance one as shown in the above example
described with reference to FIG. 9. As can be seen in FIG. 18, in
order to interpolate the fractional sample 1820 two further sample
positions 1830 would be necessary which are located out of the
window. In order to avoid additional loading of these samples from
the external memory, according to the present disclosure, the
values of the window-extern positions 1830 are replaced with values
which are obtained on the basis of the samples within the
window.
[0213] FIG. 18 illustrates fractional sample position 1820 which is
in the middle between two integer sample positions and thus a
half-pixel (half-pel) position. However, the present disclosure can
equally be applied to other fractional positions such as quarter or
1/8 positions. Moreover, sample position 1820 is located on a
horizontal line corresponding to a row of the integer sample
positions so that horizontal 1D filtering can be used for its
interpolation. However, the present disclosure is not limited to
such fractional positions. Rather, fractional positions may be
located on a vertical line corresponding to a column of the integer
sample positions so that vertical 1D filtering can be used for its
interpolation. Moreover, the fractional positions do not have to be
located on the same line of pixels as the integer samples at all.
In such case, two-dimensional filtering may be used to interpolate
such samples; such 2D filtering may be separable to vertical and
horizontal 1D filtering.
[0214] Examples of various fractional positions are shown in FIG.
15. In particular, fractional position 1 is a position for the
interpolation of which both horizontal and vertical filtering may
be applied because it is not located on line with the rows and
columns of the integer sample positions. Position 2 is a position
for the interpolation of which only horizontal filtering is used
whereas position 6 is a position for the interpolation of which
only vertical filtering is used. All fractional positions shown in
FIG. 15 can be obtained using only the integer positions within the
window.
[0215] Returning to FIG. 18, according to an embodiment, a window
of memory access for refinement is defined around position pointed
to by the non-refined motion vector 1810. The window identifies the
maximum number of pixel samples that need to be accessed from the
memory in order to perform motion vector refinement search. The
search space here correspond to the access window. In other words,
for the sake of simplicity, in this example the template size is
considered 1.times.1, but it can and in praxis would typically be
larger. The window of memory access is usually defined as the
extension around the coding block. In this example, it is 4 samples
from left/right and 4 samples from top/bottom. If search space
position currently tested during template matching requires samples
from outside of the window of memory access, then samples necessary
for reconstruction of the coding block are obtained by padding.
This is the case in FIG. 18 for the fractional position 1820. If
the template matching is to be performed on this position or using
this position, it has to be interpolated using additional padded
positions 1830 located outside the window.
[0216] Padding samples can be generated, for instance, using one of
the methods of: [0217] Nearest sample replication. [0218] Mirroring
along the axis of interpolation filtering.
[0219] In particular, the nearest sample replication refers to an
approach in which the replacement value is determined to be equal
to the value on a closest of the accessed integer sample positions.
For instance, in FIG. 18, the two missing position values 1830
would be replaced with the value of the sample immediately adjacent
to the right of the fractional position 1820 to be interpolated.
However, it is noted that this approach is only an example and the
present disclosure is not limited to a mere replication of the
neighbor. Alternatively, several nearest sample positions may be
used to interpolate the missing samples 1830. For example, three
samples on the boundary nearest to the missing sample may be used
to interpolate the missing sample for instance by weighted
averaging with weights set inverse proportional to the distance of
the three samples from the missing sample.
[0220] The mirroring refers to an approach according to which the
replacement value is determined by mirroring with respect to the
axis of the interpolation filter the value of the corresponding
assessed integer sample position. For example in FIG. 18, the two
missing samples 1830 would be replaced with the values of samples
to the left of the fractional position 1820. In particular, the
second sample to the right of the fractional position 1820 is
replaced with the value of the position second to the left of the
fractional position 1820, whereas the third sample to the right of
the fractional position 1820 is replaced with the value of the
position third to the left of the fractional position 1820.
[0221] It is noted that a combination of the two above mentioned
approaches may also be applied in which the missing samples are a
function of the mirrored samples and the nearest sample.
[0222] According to an embodiment, the window is formed by all
integer sample positions accessible for the template matching on
integer sample positions within said search space. This is for
instance the case for the example of FIG. 18, in which the
additional samples outside the window are only necessary for the
template matching in non-integer, fractional sample positions. In
other words, in FIG. 18, in the above example the half pel
coordinate that is to be searched require sample outside of the
memory access window. As a result padding operation is applied to
generate the unavailable samples. However, the present invention is
not limited thereto.
[0223] FIG. 19 shows an alternative approach in which the window
does not include all integer sample positions accessible for the
template matching on integer sample positions within said search
space. In the previous examples the coding block size was depicted
as 1.times.1 for simplicity. In FIG. 19 the size is 4.times.4
integer samples. In this example, if the search coordinate
(position on which template match is to be calculated) points to in
integer sample, padding can still be applied if some of the samples
in the block lies outside of the memory access window. This
approach enables to limit the window size even more.
[0224] In detail: FIG. 19 shows window 1920 which includes some of
the accessible samples the initial motion vector points to a
position which marks top left corner of a coding block. The search
space may be considered here as equal to the window. However, it is
noted that it can be in general smaller than the window which also
may make sense since if the template is larger than 1.times.1
integer samples, then necessarily samples apart from those included
in the search space will be necessary to perform the template
matching. Nevertheless, if the search space defines the size of the
window, alternatively, the missing samples may be padded. In FIG.
19 the exemplary search position in the second row and 10.tau.h
column of the window is shown. Template matching with a 4.times.4
integer sample large template requires also pixels outside the
window. These samples are thus padded, i.e. replaced by values
determined or calculated on the basis of the samples within the
window. The above mentioned padding approaches such as nearest
neighbor or mirroring may be applied.
[0225] The window may be defined by its vertical and/or horizontal
size with respect to [0226] the search space, or [0227] a block
with a size of the prediction block located on the initial motion
vector position, or [0228] the initial motion vector position.
[0229] An example of window definition is shown in FIG. 11 and
described above. This corresponds to determining of the window size
based on the size of the prediction block located on the initial
motion vector position. Alternatively, the window size may be
defined with respect to the initial vector position by the size in
one or both of x and y directions. For instance, the window may be
defined to be located around the position pointed to by the initial
motion vector and extending A integer samples to each of up, down,
left, right direction. Alternatively, it may extend A integer
samples to the left and right respectively and B integer samples up
and down respectively. Further definitions are possible.
[0230] A method according to an embodiment determines a motion
vector for a prediction block with the following steps illustrated
in FIG. 20. In step 2010 an initial motion vector and a template
for the prediction block is obtained, followed by determining a
refinement of the initial motion vector by template matching with
said template in a search space. The search space is located on a
position given by the initial motion vector. The memory access
window is determined in 2020. It is noted that this does not have
to be performed explicitly. The memory access window may be
predefined and known. Thus this step is for illustration to
indicate that the knowledge of the memory access window size is
known. Then the template matching is performed in steps
2030-2090.
[0231] The template matching accesses integer sample positions
within a predetermined window, the predetermined window including
said search space and further integer sample positions accessible
for the template matching; and replaces at least one integer sample
position located out of the window by a replacement value which is
based on at least one sample within said window. The accessed
integer sample position values as well as the replacement values
are used to perform the template matching. This is shown in FIG.
20: The template matching is performed for all positions in the
search space. In step 2030 it is tested whether the template
matching on the current position requires a sample or samples which
is/are out of the memory access window. If affirmative (yes in step
2030), then the padding as described above is performed for the
sample or samples out of the window in step 2040. After this step,
the template matching (i.e. search) is performed in step 2050 at
the current position using the padded samples. Alternatively, if
none of the necessary samples is out of the window (No in step
2030), the template matching (i.e. search) is performed in step
2050 at the current position using the available samples. If the
current position is the last position in the search space to be
tested (Yes in step 2060), then the template matching ends.
Otherwise (No in step 2060), then the next position is obtained in
step 2070 and the template matching continues as described above
for the new position in step 2030.
[0232] After the end of the template matching, the best matching
position is identified (e.g. as stored position with the lowest
cost).
[0233] In other words, this embodiment provides motion vector
prediction which enables to take into account the number of
accesses to the external memory and the number of samples which are
necessary to be accessible for motion vector refinement of a motion
vector for a coding block. This is achieved by replacing the values
which are located outside a predefined memory access window for the
purpose of motion vector refinement and/or fractional interpolation
with replacement values based on one or more samples from within
the memory access window. According to an aspect of the invention,
an apparatus is provided for determination of a motion vector for a
prediction block including a processing circuitry configured to:
obtain an initial motion vector and a template for the prediction
block; determine a refinement of the initial motion vector by
template matching with said template in a search space. Said search
space is located on a position given by the initial motion vector.
Moreover, said template matching: accesses integer sample positions
within a predetermined window, the predetermined window including
said search space and further integer sample positions accessible
for the template matching, replaces at least one integer sample
position located out of the window by a replacement value which is
based on at least one sample within said window, and use the
accessed integer sample position values as well as the replacement
values to perform the template matching.
[0234] Such apparatus provides an advantage of limiting the number
of samples which are to be available for the purpose of motion
vector determination (refinement) while also avoiding additional
accesses to the storage/(external) memory storing the entire
reference pictures.
[0235] In one embodiment, said search space includes at least one
fractional sample position with a value obtained by interpolation
filtering with a filter of a predefined tap-size; and the
interpolation filtering uses said accessed integer sample position
values and the at least one replacement value to obtain said at
least one fractional sample position value.
[0236] For instance, the replacement value is determined by
mirroring with respect to the axis of the interpolation filter the
value of the corresponding assessed integer sample position. The
interpolation filtering may be a one-dimensional or separable
two-dimensional filtering.
[0237] In one exemplary implementation, the window is formed by all
integer sample positions accessible for the template matching on
integer sample positions within said search space.
[0238] The window is defined, for instance, by its vertical and/or
horizontal size with respect to the search space, or a block with a
size of the prediction block located on the initial motion vector
position, or the initial motion vector position.
[0239] As another example, the replacement value is determined to
be equal to the value of a closest of the accessed integer sample
positions.
[0240] According to another aspect of the invention, an encoding
apparatus is provided for encoding video images split to prediction
blocks into a bitstream, the encoding apparatus comprising: the
apparatus for determination of a motion vector for a prediction
block as described above; and an encoding circuitry for encoding
difference between the prediction block and the predictor given by
a prediction block in a position based on the determined motion
vector and for generating bitstream including the encoded
difference and the initial motion vector.
[0241] According to another aspect of the invention, a decoding
apparatus is provided for decoding from a bitstream video images
split to prediction blocks, the decoding apparatus comprising: a
parsing unit for parsing from the bitstream an initial motion
vector and an encoded difference between a prediction block and a
predictor given by a prediction block in a position specified by a
refined motion vector; the apparatus for determination of the
refined motion vector for the prediction block as described above;
and decoding circuitry for reconstructing the prediction block as a
sum of the parsed difference and the predictor given by the
prediction block in the position specified by the refined motion
vector.
[0242] According to an aspect of the invention, a method is
provided for determination of a motion vector for a prediction
block including the steps of: obtaining an initial motion vector
and a template for the prediction block; and determining a
refinement of the initial motion vector by template matching with
said template in a search space. The search space is located on a
position given by the initial motion vector. The template matching
accesses integer sample positions within a predetermined window,
the predetermined window including said search space and further
integer sample positions accessible for the template matching;
replaces at least one integer sample position located out of the
window by a replacement value which is based on at least one sample
within said window; and uses the accessed integer sample position
values as well as the replacement values to perform the template
matching.
[0243] In one embodiment. said search space includes at least one
fractional sample position with a value obtained by interpolation
filtering with a filter of a predefined tap-size; and the
interpolation filtering uses said accessed integer sample position
values and the at least one replacement value to obtain said at
least one fractional sample position value.
[0244] For example, the replacement value is determined by
mirroring with respect to the axis of the interpolation filter the
value of the corresponding assessed integer sample position.
Moreover, in one example, the interpolation filtering is a
one-dimensional or separable two-dimensional filtering.
[0245] The window is formed, for instance, by all integer sample
positions accessible for the template matching on integer sample
positions within said search space.
[0246] According to an exemplary implementation, the window is
defined by its vertical and/or horizontal size with respect to the
search space, or a block with a size of the prediction block
located on the initial motion vector position, or the initial
motion vector position.
[0247] According to another example, the replacement value is
determined to be equal to the value on a closest of the accessed
integer sample positions.
[0248] Moreover an encoding method is provided for encoding video
images split to prediction blocks into a bitstream, the encoding
method comprising the steps of determining a motion vector for a
prediction block according to any of methods described above; as
well as encoding difference between the prediction block and the
predictor based on a prediction block in a position based on the
determined motion vector and for generating bitstream including the
encoded difference and the initial motion vector.
[0249] The encoding method may further include steps described with
reference to functions of blocks in FIG. 1.
[0250] Still further, a decoding method is provided for decoding
from a bitstream video images split to prediction blocks, the
decoding method comprising: parsing from the bitstream an initial
motion vector and an encoded difference between a prediction block
and a predictor given by a prediction block in a position specified
by a refined motion vector; determining the refined motion vector
for the prediction block according to any of methods mentioned
above; and reconstructing the prediction block as a function (such
as a sum) of the parsed difference and the predictor given by the
prediction block in the position specified by the refined motion
vector.
[0251] The decoding method may further include steps described with
reference to functions of blocks in FIG. 2.
[0252] However, it is noted that FIGS. 1 and 2 are not to limit the
present disclosure. They merely provide a non-limiting example of
an implementation of present invention within the existing encoder
and/or decoder.
[0253] The motion vector determination with initial motion vector
rounding as described above can be implemented as a part of
encoding and/or decoding of a video signal (motion picture).
However, the motion vector determination may also be used for other
purposes in image processing such as movement detection, movement
analysis, or the like without limitation to be employed for
encoding/decoding.
[0254] The motion vector determination may be implemented as an
apparatus. Such apparatus may be a combination of a software and
hardware. For example, the motion vector determination may be
performed by a chip such as a general purpose processor, or a
digital signal processor (DSP), or a field programmable gate array
(FPGA), or the like. However, the present invention is not limited
to implementation on a programmable hardware. It may be implemented
on an application-specific integrated circuit (ASIC) or by a
combination of the above mentioned hardware components.
[0255] The motion vector determination may also be implemented by
program instructions stored on a computer readable medium. The
program, when executed, causes the computer to perform the steps of
the above described methods. The computer readable medium can be
any medium on which the program is stored such as a DVD, CD, USB
(flash) drive, hard disc, server storage available via a network,
etc.
[0256] The encoder and/or decoder may be implemented in various
devices including a TV set, set top box, PC, tablet, smartphone, or
the like, i.e. any recording, coding, transcoding, decoding or
playback device. It may be a software or an app implementing the
method steps and stored/run on a processor included in an
electronic device as those mentioned above.
[0257] Summarizing, the present disclosure relates to motion vector
determination employing template matching. At first, an initial
motion vector is obtained. If the initial motion vector points to a
position which is fractional and thus, requires interpolation from
the integer sample positions, the initial motion vector is rounded
to a closest integer sample position. The rounded position is then
used to define the search space for the template matching to define
a refinement of the initial motion vector.
* * * * *
References